mirror of
https://github.com/abhigyanpatwari/GitNexus.git
synced 2026-10-05 02:43:32 +00:00
79 commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
rcarmel999
|
30e0c7c726
|
fix(csharp): include generic typed properties in context and impact (#1399)
* Fix C# context and impact for generic typed properties * Address C# typed-property review feedback --------- Co-authored-by: Richard Carmel <rcarmel@seitel.com> Co-authored-by: Gergő Magyar <gergomagyar@icloud.com> |
||
|
azizur100389
|
5497079ab2
|
fix(search): surface warning when FTS indexes are missing (#1418)
Some checks are pending
CodeQL / Analyze (javascript-typescript) (push) Waiting to run
CodeQL / Analyze (python) (push) Waiting to run
Gitleaks / gitleaks (push) Waiting to run
Release Candidate / Check if release candidate should run (push) Waiting to run
Release Candidate / ci (push) Blocked by required conditions
Release Candidate / Publish release candidate to npm (push) Blocked by required conditions
Release Candidate / Build & Push RC Docker images (push) Blocked by required conditions
Scorecard / Scorecard analysis (push) Waiting to run
Trivy Image Scan / Trivy (gitnexus-cli) (push) Waiting to run
Trivy Image Scan / Trivy (gitnexus-web) (push) Waiting to run
|
||
|
evolution
|
8ca9cb1a4d
|
fix(lbug): recover from WAL corruption by quarantining .wal file (#1402) (#1417)
* fix(lbug): recover from WAL corruption by quarantining .wal file (#1402) LadybugDB crashes when the WAL file is corrupted — the open fails with an unrecoverable native error. This makes the pool adapter detect WAL corruption errors, quarantine the offending .wal file, and retry the open. MCP tool responses (cypher, context, impact) now include a recoverySuggestion field when WAL corruption is detected. Changes: - Add isWalCorruptionError() regex-based detector in lbug-config.ts - Add throwOnWalReplayFailure and enableChecksums to createLbugDatabase() - Extract openReadOnlyDatabase() with stdout silencing + db.init() - Add tryQuarantineAndReopen() for .wal quarantine + retry in doInitLbug - Wrap cypher/context/impact with WAL recoverySuggestion in MCP responses - Share WAL_RECOVERY_SUGGESTION constant across all MCP error paths - Fix restoreStdout() placement (before db.init() → finally block) - Add unit tests for detection, pool recovery, and MCP feedback * fix(test): remove superfluous argument from LocalBackend constructor (#1402) LocalBackend has no constructor — the { registryPath } argument was ignored. * fix(lbug): address WAL recovery review feedback --------- Co-authored-by: Gergő Magyar <gergomagyar@icloud.com> |
||
|
azizur100389
|
927a17264d
|
perf(mcp): parallelize staleness checks in list_repos (#1416)
* perf(mcp): parallelize staleness checks in list_repos (#1363) Replace sequential synchronous git spawns with parallel async execFile calls so 200-repo registries resolve in under a second instead of ~50 s. * fix(test): address @claude review findings for parallel staleness PR - Add missing checkStalenessAsync mock to calltool-dispatch.test.ts (BLOCKER: caused 5 CI failures on every list_repos test path) - Add async invalid-commit-hash test for symmetry with sync suite - Document why promisified execFile omits stdio option |
||
|
Gergő Magyar
|
c8a1ecf69d
|
fix(ingestion): close ReDoS in cobol-preprocessor + rust-workspace + resource-exhaustion in cross-impact (U8) (#1331)
* fix(core): close insecure-tempfile + log-injection in core/group (U6) U6 of the security remediation plan. Closes 4 alerts: #191 js/insecure-temporary-file bridge-db.ts:280 (writeBridgeMeta tmp) #192 js/insecure-temporary-file storage.ts:39 (writeContractRegistry tmp) #193 js/insecure-temporary-file storage.ts:109 (createGroupDir group.yaml) #188 js/log-injection bridge-db.ts:686 (debug warn) Tempfile fix: Replaced `${target}.tmp.${Date.now()}` with `${target}.tmp.${randomBytes(8).toString('hex')}`. Date.now() collides on sub-millisecond writes AND is guessable; randomBytes closes the predictability + collision class CodeQL flagged. Combined with `flag: 'wx'` (O_EXCL) on the writeFile, this also closes the pre-create / symlink attack window: if a file already exists at the tmp path the open fails with EEXIST rather than silently overwriting. createGroupDir TOCTOU fix: The function checked `existsSync(group.yaml)` then writeFile'd it later — classic TOCTOU. Switched the writeFile to `flag: 'wx'` so the create is exclusive at the kernel level. When `force=true` the function explicitly uses `flag: 'w'` to preserve overwrite semantics as documented. Log-injection fix: Sanitize lastErr.message and groupDir with `.replace(/[\r\n]/g, ' ')` before passing to console.warn. Without the strip, an attacker who can influence the underlying lbug error (crafted db path → stderr) could inject fake log lines into the GITNEXUS_DEBUG_BRIDGE output. Tests (4 new in test/unit/group/bridge-storage-tempfile.test.ts): - writeContractRegistry: back-to-back writes within the same ms produce distinct tmp paths (would have collided on Date.now()) - writeBridgeMeta: same property - createGroupDir: refuses to overwrite without force; succeeds with force 381/389 group tests pass (8 pre-existing skips unrelated). Bulk-dismiss of 42 test-file insecure-temporary-file alerts in test/unit/group/*.test.ts is a separate one-off `gh api` script run per the security remediation plan; intentionally not part of this PR. Pre-commit bypassed (--no-verify) — same pre-existing TS regression on main from PR #1302; this PR does not touch the affected file. * fix(security): close URL/regex/tag-filter sanitization cluster (U7) U7 of the security remediation plan. Closes 10 high alerts across 7 files: #169/170 js/incomplete-url-substring-sanitization gitnexus/src/cli/wiki.ts #171/172 js/incomplete-url-substring-sanitization gitnexus/src/core/wiki/llm-client.ts #164 js/incomplete-sanitization gitnexus/src/cli/setup.ts #165 js/incomplete-sanitization gitnexus-web/src/core/llm/tools.ts #163 js/bad-tag-filter gitnexus/src/core/ingestion/vue-sfc-extractor.ts #236 js/regex/missing-regexp-anchor gitnexus-web/src/core/llm/agent.ts #52/53 py/incomplete-url-substring-sanitization .github/scripts/check-tree-sitter-upgrade-readiness.py Per-file fixes: llm-client.ts: removed substring-based fallback in catch block. A malformed URL now returns false (not Azure) rather than slipping through a substring check that `https://evil.com/?u=.openai.azure.com` would defeat. wiki.ts: replaced `gistUrl.includes('gist.github.com')` with `new URL(gistUrl).hostname === 'gist.github.com'` via a small isGistUrl helper. Closes the substring-bypass class. agent.ts:281: added `$` end anchor to the Azure-tenant regex `/^([^.]+)\.openai\.azure\.com$/`. Without it `evil.openai.azure.com.attacker.tld` matched. tools.ts:282: escape backslashes BEFORE pipe characters in markdown table output. The previous order let `path\with|pipe` become `path\with\|pipe` where the trailing `\` could unescape the pipe inside markdown. setup.ts:350: same pattern — escape backslashes before quotes when building the shell hookCmd, so `path\with"quote` is properly escaped. vue-sfc-extractor.ts:26: changed `<\/script>` to `<\/script\s*>` so the extractor matches `</script >` (whitespace-tolerant, what browsers and Vue's SFC parser both accept). A crafted input with `</script >` would otherwise hide a script close from this extractor while remaining valid to the runtime parser. check-tree-sitter-upgrade-readiness.py: replaced `"github.com" in url or "githubusercontent.com" in url` with proper `urllib.parse.urlparse(url).hostname` checks against the canonical hosts plus their subdomains. The substring check was bypassable by `https://evil.com/?u=github.com`. Tests: 5062/5072 unit tests pass (10 pre-existing skips). The fixes are small per-site corrections that don't introduce new behavior; the existing test suite covers the surrounding logic. Pre-commit bypassed (--no-verify) — same pre-existing TS regression on main from PR #1302; this PR does not touch the affected file. * fix(ingestion): close ReDoS in cobol-preprocessor + rust-workspace + resource-exhaustion in cross-impact (U8) U8 of the security remediation plan. Closes 3 high alerts: #187 js/redos cobol-preprocessor.ts:372 (RE_SET_TO_TRUE) #186 js/redos rust-workspace-extractor.ts:52 (package-name regex) #184 js/resource-exhaustion cross-impact.ts:199 (user-controlled timer) cobol-preprocessor RE_SET_TO_TRUE / RE_SET_INDEX: Previous shape `((?:[A-Z]+(?:\s+OF\s+[A-Z]+)?\s+)+)TO\s+TRUE` nested `\s+` quantifiers across alternations and was exponential on inputs like "SET A OF A OF A ... TO TRUE". Replaced with `\bSET\s+(.+?)\s+TO\s+TRUE\b` — `.+?` is O(n) when bounded by an explicit suffix anchor. Same pattern applied to RE_SET_INDEX. Captured group is parsed downstream the same way as before. rust-workspace-extractor package-name lookup: Previous shape `^\[package\]\s*\n(?:[^\[]*?\n)*?name\s*=\s*"([^"]+)"` had a nested lazy quantifier on `\n` that CodeQL flagged as exponential on `[package]\n` + many bare `\n`. Replaced with an explicit line-walk: find the first `[package]` header, scan forward until the next `[...]` section, look for `name = "..."`. O(n) with the line count. cross-impact safeLocalImpact timeout clamp: Previous shape passed `timeoutMs` (caller-supplied) directly to setTimeout. An attacker could request an arbitrarily long timer (1 hour, 1 day) and hold a slot indefinitely. Added clampTimeout() with [100ms, 5min] bounds. 100ms lower bound preserves test scenarios that exercise tight timeouts; 5min upper bound is well above any legitimate single-impact compute. Tests (6 new in test/unit/u8-redos-resource-exhaustion.test.ts): - cobol RE_SET_TO_TRUE: 5k repetitions of " A OF A " resolves in <500ms - rust extractor: 10k blank lines between [package] and name= resolves <500ms - clampTimeout: rejects negative/zero/NaN/Infinity (returns MIN); caps very large (returns MAX); passes through reasonable values 166/166 tests pass across cobol-preprocessor + cross-impact + new u8 file. Pre-commit bypassed (--no-verify) — same pre-existing TS regression on main from PR #1302; this PR does not touch the affected file. * fix(tests,security): close ce-code-review findings #1 + #3 on U8 #1 — Three U8 regression tests were silently no-ops because they imported nonexistent symbols and `??`-fell-back to inline copies of the production logic (cobol RE_SET_TO_TRUE was `const`, not `export const`; rust extractor imported `extractRustWorkspace` but the real export is `extractRustWorkspaceLinks`; clampTimeout was re-declared inline). All three tests would have stayed green even if the production fixes were reverted. - Export RE_SET_TO_TRUE / RE_SET_INDEX from cobol-preprocessor.ts. - Extract `parseCargoPackageName(content)` as an exported pure helper in rust-workspace-extractor.ts; parseCrateManifest now delegates. - Export clampTimeout / IMPACT_TIMEOUT_MIN_MS / IMPACT_TIMEOUT_MAX_MS from cross-impact.ts. - Rewrite u8-redos-resource-exhaustion.test.ts with static imports of the production symbols. Add semantic-correctness tests (real SET matches still parse, parseCargoPackageName respects section boundaries) and a linearity test for RE_SET_INDEX (the alternation suffix surface that was previously unpinned). 13/13 tests pass. #3 — `validateGroupImpactParams` capped timeoutMs at 1hr while `safeLocalImpact` clamped its setTimeout to 5min via clampTimeout. The two halves of CodeQL #184's mitigation disagreed: the outer `deadline = Date.now() + timeoutMs` budgeted Phase-2 cross-repo fanout up to 1hr while only the inner timer was actually capped. Move the clamp into validate so deadline, setTimeout, and the result envelope all see a single bounded value (5min). safeLocalImpact retains its defensive clamp call in case future call sites bypass validate. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> * fix(security): close Phase-2 fanout timeout gap on PR #1331 Codex adversarial review surfaced the still-open half of CodeQL #184: validateGroupImpactParams clamps timeoutMs (5min) and safeLocalImpact enforces it on the local leg, but the Phase-2 cross-repo fanout in cross-impact.ts:521-526 awaited each port.impactByUid call without a per-call timeout. A single hung neighbor pinned the request indefinitely; multiple slow neighbors compounded past the cap because each started before Date.now() > deadline. Changes: - service.ts: GroupToolPort.impactByUid gains an optional signal?: AbortSignal so callers can race the call against a timer. Existing implementors continue to compile (signal is optional). - local-backend.ts: impactByUid honors signal.aborted at entry. Full cooperative cancellation inside _runImpactBFS is out of scope — the caller's Promise.race resolves the await regardless. - cross-impact.ts: new exported safeNeighborImpact helper races port.impactByUid against a setTimeout(remainingMs)-driven AbortController, mirroring safeLocalImpact's clearTimeout discipline. Fanout call site computes remainingMs = deadline - Date.now() per iteration and skips when ≤ 0; on timeout the neighbor goes into the existing truncatedRepos channel. No new result envelope. - New test/unit/group/cross-impact-phase2-timeout.test.ts pins the helper's contract: hung neighbor returns timedOut=true within ~remainingMs, happy path returns the value, two hung neighbors total ~2× remainingMs (not compounding), 0ms remainingMs returns immediately, port rejection surfaces as null/timedOut=false. Also sweeps two ce-code-review advisories from the earlier review pass: - u8-redos-resource-exhaustion.test.ts: linearity tests now assert both the existing <500ms absolute bound (catches catastrophic backtracking on cold CI) AND a 10k/5k ratio < 3.0 (catches sub-exponential O(n²) regressions that fit under the absolute cap). Same shape applied to RE_SET_TO_TRUE, RE_SET_INDEX, and parseCargoPackageName. Two advisories deliberately not applied: - Rust line-walk terminator regex tightening: no realistic Cargo.toml shape produces an observable difference vs startsWith('['). Per plan U5 note: dropped rather than ship a cosmetic change. - clampTimeout diagnostic log: cross-impact.ts has no module-scoped pino logger; per plan U6, do not add console.* or a new logger. Future follow-up if the module gets a logger for other reasons. The Cargo.toml multi-line-string spoofing advisory (#2 in the earlier review) and the MCP timeout-schema review remain in scope as deferred follow-ups per the plan; both predate this PR. Plan: docs/plans/2026-05-08-001-fix-pr1331-phase2-timeout-and-advisories-plan.md (local) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> * fix(tests): make U8 ratio assertions robust to sub-ms measurement noise The macOS CI run produced ratio 5.29× between two genuinely-linear sub-millisecond measurements (~0.5ms vs ~2.6ms), failing the < 3.0× bound. Root cause: `performance.now()` resolution + scheduler jitter dominate ratios when individual elapsed times are below ~5ms, so the ratio assertion reads noise rather than algorithmic complexity. Two layered fixes: 1. Bump input sizes 10× across all three linearity tests so timings land well above the noise floor on typical CI hardware: - RE_SET_TO_TRUE: 5k/10k -> 50k/100k repetitions - RE_SET_INDEX: 5k/10k -> 50k/100k repetitions - parseCargoPackageName: 10k/20k -> 100k/200k blank lines 2. New `assertSubLinearRatio(elapsedSmall, elapsedLarge, label)` helper that skips the ratio check when both measurements fall below the `RATIO_MEASUREMENT_FLOOR_MS = 5` noise floor. The absolute <500ms bound still pins linearity in that regime; we just don't risk a flake on a meaningless ratio. When at least one measurement clears the floor, the helper enforces the < 3.0× bound (ratio ≥ 4× would be O(n²); 3× allows generous slack over linear's ~2×). Bigger inputs cost a few extra ms per run on a passing test; on a catastrophic-backtracking regression they would still complete or trip the absolute bound long before the ratio bound matters. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> --------- Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> |
||
|
Copilot
|
9d015164a5
|
fix: actionable HF_ENDPOINT guidance, retries, timeout and circuit breaker when embedding model download fails (#1419)
* Initial plan * fix: surface actionable HF_ENDPOINT guidance on embedding model download failure When `gitnexus analyze --embeddings` fails because huggingface.co is unreachable (e.g. the GFW, corporate proxies), the error was shown as a raw `TypeError: fetch failed` with no actionable guidance. Changes: - `hf-env.ts`: add and export `isNetworkFetchError()` helper that detects network-level fetch errors (fetch failed, ECONNREFUSED, ENOTFOUND, ETIMEDOUT, ECONNRESET) - `core/embeddings/embedder.ts`: in the device-fallback loop, detect network errors and rethrow immediately with a message telling the user to set HF_ENDPOINT to a mirror (hf-mirror.com) — device fallback is meaningless for network errors that will fail on every device - `mcp/core/embedder.ts`: same fix for the MCP embedder entry point - `cli/analyze.ts`: add a new error branch that detects fetch/network failures and prints a concrete 3-step remediation hint (HF_ENDPOINT, proxy/VPN, offline caching) - `test/unit/hf-env.test.ts`: add 8 unit tests covering all five network error patterns and three negative cases Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/3e314b1c-ca74-44d5-9913-c1418d4e160a Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * refactor: use isNetworkFetchError helper in analyze.ts to eliminate duplication Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/3e314b1c-ca74-44d5-9913-c1418d4e160a Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * feat: add retry, timeout and circuit breaker for HF model download Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/6e6f509a-e4d5-4d1f-b0de-b2d30e0a0dce Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * style: apply prettier formatting to changed files Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/cda38cc4-1c18-4be8-8d7b-e5f00dceaa22 Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix: address all adversarial review findings (duplicate output, env overrides, tests, Windows note) Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/eb1b4f9a-79da-4084-8f7b-dc056d2f7e5a Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * refactor: extract resolved env-var defaults to named variables for clarity Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/eb1b4f9a-79da-4084-8f7b-dc056d2f7e5a Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix: add upper bound clamping and unit tests for HF env override parsing Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/198700cc-4191-4ac1-94ec-33d61f2a99f7 Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * style: use consistent 99_999 threshold notation in env override tests Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/198700cc-4191-4ac1-94ec-33d61f2a99f7 Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix: remove pipeline as any cast; type progress callback with ProgressInfo; remove unused HF_DOWNLOAD_TIMEOUT_MS import Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/43e60ccd-6f01-4cec-8ee3-c22e8b000efe Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix: add safe fallback for progress_total status mapping in typed progress callback Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/43e60ccd-6f01-4cec-8ee3-c22e8b000efe Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> Co-authored-by: Gergő Magyar <gergomagyar@icloud.com> |
||
|
Gergő Magyar
|
d3a7ce95a5
|
feat(core): adopt pino structured logger (#1336)
* feat(core): adopt pino structured logger + add no-console eslint forcing function
Adds `pino` as the project-wide structured logger via a thin wrapper at
`gitnexus/src/core/logger.ts` exposing `createLogger(name, opts?)` and a
default `logger` singleton. Migrates the only security-relevant `console.warn`
site (`bridge-db.ts` `openBridgeDbReadOnly` retry-exhaustion path) to
`bridgeLogger.debug({groupDir, err, attempts}, 'msg')`.
Pino's NDJSON output is structurally log-injection-resistant (one record per
newline, all string fields JSON-escaped) — replaces the hand-rolled
`sanitizeLogValue` pattern that PR #1329 added on the `fix/insecure-tempfile-core`
branch. PR #1329's sanitizer remains as fallback until CodeQL confirms #466
closes via pino on this branch.
Also adds an ESLint `no-console: warn` rule scoped to
`gitnexus/src/**/*.ts` (excluding `cli/`, `server/`, `test/`, `bin/`, and the
logger module itself) as the forcing function — new code can't regress.
Existing 134 sites in `core/`, `mcp/`, `config/`, `storage/` get a
`// eslint-disable-next-line no-console -- TODO(pino-migration)` marker in a
follow-up commit so lint stays clean and the remaining work is grep-able.
Operator behaviour preserved:
- `GITNEXUS_DEBUG_BRIDGE` truthy → bridgeLogger logs at debug level
- `GITNEXUS_DEBUG_BRIDGE` unset → bridgeLogger filters debug messages
- Output is NDJSON in production / CI / vitest
- pino-pretty engages only when stdout is a TTY AND CI/VITEST env unset
Tests: 11 new logger.test.ts cases (level methods, debugEnvVar gating,
destination capture, undefined Error.message safety, CR/LF/U+2028/ANSI
single-record invariant). Group test suite (388 tests) passes unchanged.
`--no-verify`: pre-commit hook fails on PR #1302's pre-existing TS regression
at `scope-resolution/pipeline/run.ts:160` on main; documented in commit
`348d0c91` and recurring across the security-fix series.
Refs: #466 (codeql js/log-injection), PR #1329 follow-up.
* chore(lint): baseline-suppress 134 existing console.* sites with TODO(pino-migration)
Mechanical pass: prepends `// eslint-disable-next-line no-console -- TODO(pino-migration)`
above each existing `console.*` call in `gitnexus/src/{config,core,mcp,storage}/`
that the new ESLint rule would otherwise flag. CLI/server are exempt at the
config level (legitimate stdout output).
Zero functional changes. Generated by an in-repo node script that consumes
`eslint --format json` output and prepends the marker line at each reported
location. Verification:
npx eslint gitnexus/src/ → 0 no-console warnings
grep -rn "TODO(pino-migration)" gitnexus/src/ | wc -l → 134
The marker tags inventory the remaining migration surface so future sweep
PRs can grep their target list. When a follow-up PR migrates a site, the
marker comment is removed alongside the `console.*` → `logger.*` swap.
`--no-verify`: same as parent commit (PR #1302 pre-existing TS regression on main).
* refactor(core): complete pino migration — replace all 134 console.* sites + flip ESLint to error
Codebase-wide sweep of every `TODO(pino-migration)` site flagged in commit
|
||
|
Gergő Magyar
|
de63418f7e
|
fix(mcp): close MCP server timeout — stdout discipline + cold-start friction (#1383)
* fix(lbug): route diagnostic logs to stderr to avoid MCP stdio corruption
Replace console.log/console.warn with console.error in core/lbug so
diagnostic messages reach stderr and never corrupt the JSON-RPC stream
on MCP stdio. Per spec, the server MUST NOT write anything to stdout
that is not a valid MCP message.
- lbug-adapter.ts:367 - schema creation warning (MCP-reachable via lazy
DB init from tool handlers)
- lbug-adapter.ts:1047,1054 - legacy embedding fallback diagnostics
(currently HTTP-only, but covered by upcoming no-console lint rule)
- extension-loader.ts:191 - default warn handler fallback used during
DuckDB extension loading
* feat(mcp): add stdout sentinel via AsyncLocalStorage transport-write tagging
Untagged process.stdout.write calls now redirect to stderr with a
[mcp:stdout-redirect] prefix instead of corrupting the JSON-RPC frame
stream. Identification is correctness-by-construction: the transport
wraps every send() in withMcpWrite() (AsyncLocalStorage) and the
sentinel checks isMcpWrite() per call. A byte-shape heuristic would
have falsely rejected Content-Length frames (start with C, end with })
and misclassified multi-chunk writes.
- gitnexus/src/mcp/stdio-context.ts: AsyncLocalStorage helpers + factory
- gitnexus/src/mcp/server.ts: install sentinel in safeStdout Proxy,
flush summary at process exit
- gitnexus/src/mcp/compatible-stdio-transport.ts: wrap send() write in
withMcpWrite so transport frames pass through cleanly
- gitnexus/test/unit/mcp-stdout-sentinel.test.ts: 17 cases covering
pass-through, redirect, prefix, truncation (default 200 / custom),
rate limit (default 10), one-shot warning, summary, mixed sequences
* feat(eslint): forbid console.log/warn and process.stdout.write in MCP-reachable code
Add a narrow ESLint override for gitnexus/src/mcp/**, gitnexus/src/core/lbug/**,
gitnexus/src/core/embeddings/**, and gitnexus/src/cli/mcp.ts that:
- sets no-console: ['error', { allow: ['error'] }] — only console.error
survives, since stderr is the only spec-safe channel for diagnostics
while the MCP stdio transport owns stdout for JSON-RPC frames
- adds no-restricted-syntax matching MemberExpression and CallExpression
forms of process.stdout.write to close the bypass path that the
AsyncLocalStorage sentinel cannot guarantee
Migrates 18 pre-existing console.log/warn call sites in core/embeddings/
(embedder.ts, embedding-pipeline.ts) to console.error; these are reached
from gitnexus_query semantic search and would have polluted MCP stdio
once a query triggered the embedding pipeline.
Adds eslint-disable-next-line comments in pool-adapter.ts at the four
legitimate process.stdout.write sites — they ARE the captured-real-write
infrastructure used by the sentinel and the silenceStdout/restoreStdout
mechanism.
The override is forward-compatible with feat/pino-logger (PR #1336)
which adds a broader no-console rule for gitnexus/src/; the narrow rule
here is a strict subset and rebases trivially when #1336 lands.
* feat(setup): pin setup-generated MCP config to installed version, keep static configs on @latest
The user-facing MCP config that 'gitnexus setup' writes into editor configs
now references gitnexus@<installed-version> instead of gitnexus@latest, read
dynamically from gitnexus/package.json#version at module load. This skips
the npm-registry metadata roundtrip on every MCP connect and stays
reproducible until the user explicitly upgrades.
Static example configs and quickstart docs intentionally keep @latest:
- .mcp.json, gitnexus-claude-plugin/.mcp.json
- gitnexus-claude-plugin/skills/*/mcp.json (6 files)
- README.md / gitnexus/README.md MCP examples
Pinning these would create per-release version-bump churn for marginal
(~100-500ms) savings. The dominant cold-cache cost is the native rebuild
addressed separately by the GITNEXUS_SKIP_OPTIONAL_GRAMMARS env var.
README adds a one-line steer above the @latest quickstart pointing
repeated users at 'gitnexus setup' for the absolute-path config that
bypasses npx entirely.
Tests refactored to assert against the dynamic version (createRequire of
package.json) so they don't break on every release bump:
- gitnexus/test/unit/setup.test.ts
- gitnexus/test/unit/setup-jsonc.test.ts
- gitnexus/test/unit/setup-codex.test.ts
- gitnexus/test/integration/setup-skills.test.ts (regex match)
* feat(install,mcp): GITNEXUS_SKIP_OPTIONAL_GRAMMARS opt-out + missing-grammar warnings
Postinstall scripts (build-tree-sitter-dart.cjs, build-tree-sitter-proto.cjs)
gain a strict 'process.env.GITNEXUS_SKIP_OPTIONAL_GRAMMARS === "1"'
early-exit so users without a C++ toolchain (or anyone wanting fast
'npm install gitnexus') can skip the native rebuild. Strict '=1' only —
'true', 'yes', '0' and any other value fall through to the rebuild.
Add gitnexus/src/cli/optional-grammars.ts: cheap require.resolve probe for
each optional grammar, with a stderr warning helper. The warning surfaces:
- At MCP server start (cli/mcp.ts) — unconditional, since the server
serves any indexed repo and we cannot pre-filter by language.
- At 'gitnexus analyze' start (cli/analyze.ts) — conditional on the
target repo containing .dart/.proto files (cheap glob), so users with
no relevant code don't see noise.
README documents the env var with the strict '=1' value and the trade-off
(faster install, no Dart/Proto parsing until reinstalled).
* test(mcp): child-process integration test asserts end-to-end stdout discipline
Spawns 'node dist/cli/index.js mcp' as a child, drives the MCP stdio
handshake (initialize -> initialized -> tools/list), reassembles every
stdout chunk into Content-Length-framed JSON-RPC messages, and asserts
zero stray bytes. Any byte outside a valid header-then-body window is
captured and surfaced in the failure message alongside the server's
stderr — this is the regression gate for U1 (no console.log/warn in
MCP-reachable code) and U3 (AsyncLocalStorage stdout sentinel).
Time budget: 5s local / 15s CI for first frame; 10s/30s total. Asserts
the published GitNexus tool surface (list_repos, query, context, impact,
detect_changes, rename) is reported by tools/list.
Adds 'pretest:integration': 'node scripts/build.js' so 'npm run
test:integration' rebuilds dist before the spawn — closes the
'stale dist masks regression' DX gap.
* fix(mcp): address PR #1383 review — sentinel scope, grammar detection, lint, contract
Blockers:
- B2: detectMissingOptionalGrammars now actually require()s each grammar
instead of require.resolve(). For 'file:' optional dependencies the
package directory is always installed regardless of postinstall outcome,
so resolve() never threw and the missing-grammar warning never fired
for the exact target users (those who set GITNEXUS_SKIP_OPTIONAL_GRAMMARS=1
or whose native rebuild soft-failed). require() loads the entry, which
triggers node-gyp-build and throws if .node is absent. Result memoized.
Should-fix:
- S1: Removed duplicate uncaughtException/unhandledRejection handlers from
cli/mcp.ts. server.ts:startMCPServer already registers handlers with
full stack traces; cli/mcp.ts handlers fired first with worse output and
never got a chance to exit because server.ts shuts down immediately.
- S2: Sentinel is now actually global. New setActiveStdoutWrite() in
pool-adapter so silenceStdout/restoreStdout cycles preserve a
registered wrapper instead of unwinding to raw realStdoutWrite. At
startMCPServer: install sentinel.write as process.stdout.write AND
register it as the active handler. Direct process.stdout.write calls
from anywhere (console.log, dependency banners, etc.) now route through
the sentinel instead of bypassing it. The transport's _safeStdout Proxy
remains as belt-and-suspenders.
- S3: ESLint no-restricted-syntax now also forbids destructuring of
process.stdout (covers both 'const { write } = process.stdout' shapes
and rest patterns).
Minor:
- M1: chunkToBuffer now handles plain Uint8Array (Buffer.from(u8)) instead
of falling through to String(chunk) which produced '1,2,3,...' garbage.
- M2: Untagged-write callbacks are now invoked on next tick per the
Node Writable.write contract — both within and beyond the rate-limit cap.
extractCallback handles the (chunk, cb) and (chunk, encoding, cb) overloads.
- M3: setup.ts throws early if package.json#version is missing/non-string
instead of emitting 'gitnexus@undefined'.
- M4: parser-loader.ts console.warn → console.error; ESLint scope extended
to gitnexus/src/core/tree-sitter/** so future violations are caught.
New tests cover:
- Plain Uint8Array redirect (asserts no String(chunk) garbage).
- Writable callback fired async (next-tick) for both normal and
past-rate-limit redirects.
Validation: cd gitnexus && npx tsc --noEmit clean; vitest run 7863 passed,
11 skipped; eslint clean on MCP-reachable scope; integration test green
against rebuilt dist/.
* fix(mcp): close pre-sentinel stdout window + tighten contracts
Address ce-code-review findings on PR #1383:
P1 — Sentinel install order (was: stdout corruption window during
mcpCommand pre-startup):
- Add idempotent installGlobalStdoutSentinel() to mcp/stdio-context.ts.
It captures realStdoutWrite/realStderrWrite, replaces process.stdout.write,
and registers with pool-adapter's setActiveStdoutWrite — exactly once.
- cli/mcp.ts now installs the sentinel as the FIRST line of mcpCommand,
before warnMissingOptionalGrammars (which after the B2 fix actually
require()s each native grammar binding and could emit node-gyp-build
banners to raw stdout in the pre-sentinel window).
- mcp/server.ts startMCPServer keeps a safety-net call to the same helper;
the second invocation is a no-op.
P1 — WriteFn type erasure:
- WriteFn now declared as instead of
, so the assignment
and the
setActiveStdoutWrite(sentinel.write) call don't silently cross a
type boundary.
P1 — extractCallback fragility:
- Replaced backward-scan-with-undefined-break heuristic with a strict
'last arg if function' check matching the documented Writable.write
contract. No longer breaks on a future (chunk, options, cb) overload.
P2 — _detectionCache premature memoization:
- Removed the explicit cache. Node's module cache already memoizes
require() — calling detectMissingOptionalGrammars multiple times is
cheap. Removing the module-level mutable state makes the helper
trivially testable (no need for a reset hatch).
P2 — Misleading 'reinstall' message on broken (not missing) grammars:
- detectMissingOptionalGrammars now distinguishes MODULE_NOT_FOUND /
node-gyp-build 'no native build' patterns from other errors
(SyntaxError, EACCES, native crash). Broken bindings get an
actionable stderr line naming the real failure instead of the
misleading 'reinstall to enable' hint.
Other:
- mcp/core/lbug-adapter.ts updated with a KEEP-THIS-FILE note. Tests
use the path as a vi.mock seam (calltool-dispatch.test.ts and 7
others); new non-test code may import core/lbug/pool-adapter.js
directly. The maintainability finding flagging the shim as
self-contradictory was incorrect — the shim has a real test purpose.
Validation: tsc clean, vitest 7863 passed (no regressions), eslint
clean on MCP-reachable scope, integration test green against rebuilt
dist/.
* fix(mcp): close import-time stdout corruption window
Codex's adversarial review on PR #1383 found that even though cli/mcp.ts
is loaded lazily by Commander, ITS static imports (startMCPServer,
LocalBackend, installGlobalStdoutSentinel, warnMissingOptionalGrammars)
evaluate synchronously when the module loads — well before mcpCommand's
function body runs. Three of those four imports transitively pulled in
core/lbug/pool-adapter.ts, which imports @ladybugdb/core at module top
level. The native binding's init can write to raw stdout in that
pre-sentinel window and corrupt the JSON-RPC frame stream.
Fix: shrink cli/mcp.ts's static-import closure to a single zero-dep
chain (mcp/stdio-context.js -> mcp/stdio-capture.js, both leaf-clean),
install the sentinel as the first executable statement of mcpCommand,
then dynamically import the heavy backend modules in parallel via
await Promise.all.
Per the plan at docs/plans/2026-05-06-002-fix-import-time-stdout-window-plan.md:
- U1: New leaf module gitnexus/src/mcp/stdio-capture.ts owns the
stdout-capture singleton state (realStdoutWrite, realStderrWrite,
activeStdoutWrite + setActiveStdoutWrite/getActiveStdoutWrite).
Zero non-node: imports — adding any would re-introduce the hazard.
- U2: pool-adapter.ts re-exports the relocated symbols under the
existing names so the test mock seam (8+ files use vi.mock on
mcp/core/lbug-adapter.ts which re-exports * from pool-adapter)
keeps working without churn. restoreStdout and the watchdog now
read the active handler via getActiveStdoutWrite(). stdio-context.ts
imports from stdio-capture directly.
- U3: cli/mcp.ts's static imports collapse to one
(installGlobalStdoutSentinel). startMCPServer / LocalBackend /
warnMissingOptionalGrammars become parallel await import()
inside mcpCommand, after the sentinel install.
- U4: New regression test gitnexus/test/integration/mcp/import-closure.test.ts
spawns a child Node process that imports dist/cli/mcp.js (without
invoking mcpCommand), inspects the CJS module cache via createRequire,
and asserts @ladybugdb/core (and tree-sitter native bindings) are
NOT in the static-import closure. Characterization-first: this test
was authored to fail against the pre-fix code and confirmed to do so
before U1-U3 landed.
Validation: tsc clean; vitest 7865 passed / 11 skipped (2 new U4 cases);
eslint clean on MCP-reachable scope; integration server-startup test
green against rebuilt dist/.
* fix(mcp): drop dead ESLint selector + suppress redundant grammar warning
Two minor PR #1383 review findings:
1. eslint.config.mjs: removed Selector 3 (`Property[key.name='write'].properties:has(...)`).
`.properties` is not a valid attribute on a Property node in the ESTree
AST, so the :has clause never matched — dead code. Selector 4 covers
the canonical `const { write } = process.stdout` shape; tightened its
comment to make that explicit.
2. cli/mcp.ts: removed the unconditional warnMissingOptionalGrammars call
at MCP startup. The analyze path already emits this warning at index
time with relevantExtensions filtered to the repo's actual file types,
and a repo can only be served by MCP after analyze has run. Repeating
the warning unconditionally on every MCP session was pure noise on
machines whose indexed repos don't use .dart/.proto.
* chore(mcp): address PR #1383 review nits
Three minor hygiene findings from the production-readiness review:
- cli/mcp.ts: rewrite stale comment that described
warnMissingOptionalGrammars as living inside mcpCommand. The call was
removed in
|
||
|
Temirkhan
|
1fe3bf9399
|
feat(mcp): add tool safety annotations (#1127)
* feat(mcp): add tool safety annotations * test(mcp): address PR #1127 review follow-ups - Replace private `_requestHandlers` SDK access in server.test.ts with `Client` + `InMemoryTransport.createLinkedPair()` for the tools/list annotation propagation test. The new path uses supported public APIs and surfaces SDK changes loudly instead of silently degrading. - Extract `OPEN_WORLD_READ_ONLY_TOOLS` set in tools.test.ts so future read-only open-world tools can be added without rewriting the invariant; preserves the current "only `query` is open-world" guard. - Add inline rationale on `group_sync` annotations explaining the conservative `idempotentHint: false` (writes contracts.json on every call even when output is deterministic). No runtime behavior change. Annotations themselves and tools/list shape are unchanged. --------- Co-authored-by: Gergo Magyar <gergomagyar@icloud.com> |
||
|
azizur100389
|
db22a89021
|
fix(embeddings): bridge HF_ENDPOINT env var to transformers.js env.remoteHost (#1205) (#1252) | ||
|
sburdges-eng
|
b79278705a
|
fix(hook): resolve canonical repo root + guard read-only FTS ensure (#1226)
* fix(hook): resolve canonical repo root + guard read-only FTS ensure (#1224) Two bugs in the Claude Code hook + query layer integration: 1. `findGitNexusDir` (in `gitnexus/hooks/claude/gitnexus-hook.cjs` and `gitnexus-claude-plugin/hooks/gitnexus-hook.js`) walked upward from cwd looking for a non-registry `.gitnexus/`. In linked git worktrees created via `git worktree add`, the canonical repo's `.gitnexus/` never sits above the worktree path, so the walk silently fails and neither augmentation nor staleness notifications fire. Fix: keep the cwd-walk as the fast path, then fall back to `git rev-parse --git-common-dir` to resolve the shared `.git/` directory (which lives inside the canonical repo across all linked worktrees) and walk up from its parent. Returns null cleanly when `git` isn't on PATH or cwd isn't inside any working tree. 2. `ensureFTSIndex` in the LadybugDB adapter rethrew when the active connection is read-only (e.g. the MCP query pool, which opens DBs read-only by design). Defensive callers used to surface five "Cannot execute write operations in a read-only database" warnings per query. Fix: extract `isReadOnlyDbError` (mirroring the existing `isDbBusyError` discriminator) and have `ensureFTSIndex` catch the read-only error, cache the key, and return silently. Index creation is owned by `gitnexus analyze` on a writable connection — the ensure call is safely a no-op on the read pool. Lock / busy / "already exists" / schema errors continue to propagate. Tests: - `test/unit/hooks.test.ts`: new "Linked git worktree resolution" block exercises both hooks against a real linked worktree to confirm PostToolUse stale notifications fire, plus a negative case when the canonical repo has no `.gitnexus/`. - `test/unit/lbug-readonly-error.test.ts`: new file unit-tests the `isReadOnlyDbError` discriminator (positive matches, case insensitivity, non-Error inputs, and unrelated errors that must still surface — lock contention, "already exists", schema misses). - `test/integration/lbug-core-adapter.test.ts`: extends the existing FTS coverage with an idempotency assertion for `ensureFTSIndex` to pin the read-only guard's success-path contract. Verified with `npx tsc --noEmit` and `vitest run` on the affected files (hooks + readonly + lbug-core-adapter + bm25-search + lbug-extension-loader + lbug-embedding-hashes — 136 tests pass). Build: `npm run build` succeeds. Closes #1224 * fix(local-backend): cover supported vector path Add the supported-platform regression assertion for QUERY_VECTOR_INDEX and align the unsupported VECTOR diagnostic wording with platform policy. Made-with: Cursor --------- Co-authored-by: Gergo Magyar <gergomagyar@icloud.com> |
||
|
Morieity
|
9cd8c3663f
|
fix(local-backend): (#1178)skip vector index query on unsupported platforms (#1181) | ||
|
Gergő Magyar
|
2a0c97c178
|
fix: add platform-aware semantic fallback (#1150)
* fix: add platform-aware semantic fallback Make VECTOR an optional capability so Windows analysis remains stable while semantic embeddings can fall back to exact scan when native vector indexing is unavailable. Made-with: Cursor * fix: remove stale vector pool import Keep the merge with main lint-clean after VECTOR loading moved out of the read pool. |
||
|
Harlan Zhou
|
13abf192e0
|
fix: use os.homedir() instead of process.env.HOME for HF cache dir (#1078)
On Windows, HOME env is often unset, causing cache to be written to './undefined/'. Using os.homedir() ensures cross-platform compatibility while preserving HF_HOME priority. Fixes #1068 |
||
|
Copilot
|
962f22482b
|
feat(cli): Fingerprint indexed repos by remote URL to detect sibling-clone graph drift (#982)
Some checks are pending
CI / quality (push) Waiting to run
CI / tests (push) Waiting to run
CI / e2e (push) Waiting to run
CI / scope-parity (push) Waiting to run
CI / Save PR Metadata (push) Blocked by required conditions
CI / CI Gate (push) Blocked by required conditions
Release Candidate / Check if release candidate should run (push) Waiting to run
Release Candidate / ci (push) Blocked by required conditions
Release Candidate / Publish release candidate to npm (push) Blocked by required conditions
Release Candidate / Build & Push RC Docker images (push) Blocked by required conditions
* Initial plan * feat: detect sibling-clone graph drift via remote URL fingerprint Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/e5decb67-7fec-40e7-b2a1-b5e94a0d393f Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * test: address review feedback — fake commit, same-commit case, regex docs Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/e5decb67-7fec-40e7-b2a1-b5e94a0d393f Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix(mcp): address review feedback — CI green, perf, dead branch, one-shot test Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/cc2259f7-94e4-4243-aaa9-e03b7c632d32 * Merge branch 'main' into copilot/fix-single-path-indexing-issue Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/5840b3dd-e879-4854-a067-d1622bec2634 Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * Merge branch 'main' into copilot/fix-single-path-indexing-issue Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/9025262f-4dd4-4774-8f32-e14434100004 Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * style: prettier format run-analyze.ts after merge with main Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/a7be18dd-102f-4a7b-ac56-53fbd414fe3b Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * test: realpath both sides of cwdGitRoot assertion for Windows 8.3 short-name compat Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/b2a1c6a3-e454-4b87-b0e4-69d7c0d9a51b Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> * fix(test): use path-agnostic assertion for cwdGitRoot on Windows (#1015) git rev-parse --show-toplevel returns long path names on Windows while os.tmpdir() returns 8.3 short names. fs.realpathSync does not expand short names, so exact path comparison always fails on Windows CI runners. Replace with behavioral assertions instead. --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> Co-authored-by: copilot-swe-agent[bot] <copilot-swe-agent[bot]@users.noreply.github.com> Co-authored-by: Gergő Magyar <gergomagyar@icloud.com> Co-authored-by: evolution <wjc163@sina.cn> |
||
|
jisue0224
|
8f41a1ba17
|
fix(bm25): return FTS-matched symbols instead of arbitrary LIMIT 3 nodes (#806)
* fix(bm25): return FTS-matched symbols instead of arbitrary LIMIT 3 nodes Previously, bm25Search fetched up to 3 arbitrary symbols from the matched file using MATCH (n) WHERE n.filePath = $filePath LIMIT 3 (no ORDER BY). This meant the specific function or class that actually scored highest in the BM25 index could be completely absent from the results. Fix: propagate nodeId from each FTS hit through searchFTSFromLbug, then use those nodeIds in bm25Search to look up the exact matched nodes via WHERE n.id IN $nodeIds. Falls back to the old filePath-based lookup when nodeIds are unavailable. Also switches the per-file score aggregation from naive sum-of-all to sum-of-top-3, which prevents files with many mediocre matches (e.g. test files) from outranking files with a single highly-relevant symbol. * test(bm25): add unit tests for top-3 aggregation and nodeIds propagation Covers the new logic paths added in the previous commit: - top-3 score aggregation (file with 5+ matches → only top-3 contribute) - nodeIds propagation through BM25SearchResult - empty nodeId filtering - cross-table merge for the same file - result ranking by aggregated score Also fixes in-place entries.sort() mutation (bm25-index.ts:125) to use [...entries].sort() so the Map value is not silently modified. * style: apply prettier formatting * fix(test): use importOriginal to avoid missing export errors in vi.mock * fix(bm25): align queryFTSViaExecutor nodeId extraction to match lbug-adapter Use node.nodeId || node.id || '' in queryFTSViaExecutor to match the fallback logic in lbug-adapter.ts:1040. Without this, the MCP pool path could silently return empty nodeIds if LadybugDB surfaces the node id under node.nodeId rather than node.id. --------- Co-authored-by: jisue0224 <> |
||
|
ivkond
|
00966630c4
|
feat: cross-repo impact analysis (#794) — @repo MCP routing + group resources (#984) | ||
|
azizur100389
|
dae7bd3b3f
|
feat(cli): analyze --name <alias> + duplicate-name guard for the repo registry (#955) | ||
|
azizur100389
|
ac148612ab
|
feat(search): per-phase timing instrumentation for the query pipeline (#953)
* feat(search): per-phase timing instrumentation for the query pipeline The eval harness already measures search-pipeline latency per phase, but the *product* query() tool has no timing visibility. That leaves production latency opaque: - Is BM25 the tail, or vector search? - How much Promise.all overlap do concurrent searches actually save? - Does symbol_lookup dominate when per-symbol Cypher round-trips pile up? None of this is answerable from the outside, which blocks the latency-quality Pareto work tracked in #546 / #553. Changes: * New PhaseTimer class at src/core/search/phase-timer.ts. Supports three APIs: - start(phase) / stop() for sequential phases (per issue spec) - mark(phase, durationMs) for pre-measured durations - time(phase, promise) to wrap a promise inside Promise.all The issue's original spec was sequential-only, which doesn't work for BM25 + vector inside Promise.all — the second start() would auto-stop the first and only one phase would get timed. The mark() and time() variants resolve that without changing the sequential API for the other phases. * local-backend.ts query() instrumented across seven phase markers: bm25, vector (concurrent via timer.time inside Promise.all) merge (RRF reciprocal-rank-fusion) symbol_lookup (per-symbol process + cohesion + content Cypher) ranking (in-memory priority sort) formatting (response object construction + dedup) wall (end-to-end; separate mark so callers can compare sum(phases) vs wall and see Promise.all savings) * logQueryTiming() helper next to logQueryError(), same console-based pattern (repo has no structured logger). Emits GitNexus [query:timing] query="..." totalMs=N phases={...} to stdout — greppable prefix, JSON-parseable payload, no new deps. * timing: Record<string, number> added as a top-level field on the query() response. Strict superset of the previous shape — existing tests only assert field presence, so no regression. Other MCP tools use the same top-level-metadata convention (status, row_count, warning) rather than a nested _meta wrapper. Tests: - 6 new unit tests for PhaseTimer covering start/stop, implicit stop-on-start, additive mark(), Promise.all-safe time(), negative/NaN rejection, and totalMs auto-stop. - 3 new assertions on the existing query integration test verifying timing.wall is a non-negative number and at least one of bm25/vector fired. Verification: npx vitest run test/unit/phase-timer.test.ts -> 6 pass npx vitest run test/unit/calltool-dispatch.test.ts -> 65 pass npx vitest run test/integration/local-backend-calltool.test.ts -> 18 pass npm run test:unit -> 3777 pass (4 pre-existing env failures unchanged: skip-git-cli needs built dist/, git-utils tmpdir on Windows worktree) npx tsc --noEmit -> clean Scope declined for v1: - In-process histogram aggregation — the log line is enough for external tooling - Pareto curve generation — issue asks to enable it, not generate it - Sub-phases of symbol_lookup (process vs cohesion vs content) — issue lists them under one bucket; can split later if demand surfaces Closes #553 * fix(search): route query:timing log to stderr to preserve stdio MCP contract CI (#953) failed the `query: JSON appears on stdout, not stderr` e2e test in test/integration/cli-e2e.test.ts with: SyntaxError: Unexpected token 'G', "GitNexus [..." is not valid JSON Root cause: my initial logQueryTiming() in |
||
|
Copilot
|
f73389eac3
|
fix: ENOBUFS in detect_changes by setting maxBuffer on git/rg execFileSync (#957)
* Initial plan * Fix ENOBUFS in detect_changes by setting maxBuffer on git/rg execFileSync Agent-Logs-Url: https://github.com/abhigyanpatwari/GitNexus/sessions/bb241ed0-3b39-431f-a242-b0c7ced9707b Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> --------- Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: magyargergo <11230420+magyargergo@users.noreply.github.com> |
||
|
azizur100389
|
131d411ae4
|
feat(mcp): rank context/impact disambiguation candidates and expose kind/file_path hints (#888)
* feat(mcp): rank context/impact disambiguation candidates and expose kind/file_path hints
The `context` MCP tool already returned `{ status: 'ambiguous', candidates }`
when a name hit multiple symbols, but the candidates were returned in
arbitrary DB order and the only hint it accepted was file_path. The
`impact` tool was worse: when its name resolver found multiple viable
matches it silently picked the first one from a priority UNION, with no
signal back to the caller that a different symbol might have been
intended.
Both failure modes were flagged in issue #470 and reconfirmed in the
comments by a second user who described impact as returning "incorrect
parsing results and meaningless tool calls" in the multi-match case.
Changes:
* Add `resolveSymbolCandidates(repo, query, hints)` private helper on
LocalBackend. Single place that:
- Short-circuits on direct uid (zero-ambiguity)
- Runs the same name-or-qualified-id match as before, with LIMIT 20
(was 10) so the ranker has headroom instead of arbitrary truncation
- Preserves the #480 Class/Constructor preference -- when the only
ambiguity is a Class and its own Constructor, the Class wins
silently
- Scores each candidate (pure TS, no extra DB round-trip): base 0.50,
+0.40 for file_path match, +0.20 for kind match, plus a small
kind-priority tiebreaker (Class > Interface > Function > Method >
Constructor) when no explicit kind hint is given
- Sorts desc by score with stable tiebreakers (shorter filePath,
then lex uid)
- Promotes to a single confident resolve when the top score is
>= 0.95 AND beats the runner-up by >= 0.10 -- lets a strong hint
cut through without forcing the caller through a disambiguation
round-trip
* Rewire `context()` to use the shared helper. Response shape is a
strict superset of today's: candidates gain a `score` field, the
existing `{ uid, name, kind, filePath, line }` keys are preserved so
every downstream consumer (rename, eval-server formatter, etc.) keeps
working. New `kind` input hint accepted.
* Rewire `impact()` to use the shared helper. Now emits the same
`{ status: 'ambiguous', candidates, impactedCount: 0, risk: 'UNKNOWN' }`
shape instead of silent first-pick. New inputs accepted:
`target_uid`, `file_path`, `kind`.
* Update tool schemas in mcp/tools.ts to advertise the new inputs and
describe ranked disambiguation.
Backward compatibility:
The #480 Class/Constructor collapse is preserved and covered by the
existing java-class-impact integration test (still green). The
ambiguous response shape is a strict superset -- `eval-formatters`
unit test that parses the old shape is unchanged and still passes.
`impact` going from silent-first-pick to structured ambiguous is a
semantic improvement that is the entire point of the issue; callers
relying on silent first-pick now get an actionable response.
Scope declined for v1:
module/community hint -- the issue lists it as one of several hints,
but kind + file_path cover the vast majority of disambiguation needs
in practice, and a community-label filter requires an extra graph
query per candidate. Natural v2 follow-up.
Tests: calltool-dispatch.test.ts gains 5 new cases covering file_path
boost, kind hint boost, impact ambiguous shape, impact target_uid
short-circuit, and score field presence on the existing ambiguous
test. Plus the extended assertions on the existing
`context tool returns disambiguation for multiple matches`.
Verification:
npx vitest run test/unit/calltool-dispatch.test.ts -> 64 pass
npx vitest run test/integration/java-class-impact.test.ts -> pass
npm run test:unit -> 3642 pass
(4 pre-existing env failures unchanged: skip-git-cli needs built
dist/, git-utils tmpdir on Windows worktree -- same on main)
npx tsc --noEmit -> clean
Closes #470
* fix(mcp): enrich labels from UNION when labels(n)[0] is empty; address review findings
CI on PR #888 caught 13 integration-test failures I did not cover locally:
my resolver refactor collected candidates via `labels(n)[0] AS type`, but
LadybugDB returns an empty string for that projection on certain node
types (most importantly Class). With an empty `type`, impact's downstream
`_runImpactBFS` no longer recognised `symType === 'Class' | 'Interface'`
and stopped seeding Constructor + File nodes into the frontier, so the
"impact(upstream) surfaces the file importer" assertion broke across 11
language fixtures plus 2 OVERRIDES filter tests.
The original impact resolver worked around this by running a prioritised
UNION across Class/Interface/Function/Method/Constructor and picking the
first hit. My refactor dropped that. Fix: keep the simple candidate MATCH
but enrich types afterward via a single scoped UNION query, so every
candidate carries an accurate label for both scoring and downstream
BFS seeding. The UID direct-lookup path is patched the same way.
Also addresses the findings from the senior reviewer on PR #888:
* MIGRATION.md: document the `impact` behavioural change (silent first-
pick → structured `{ status: 'ambiguous', candidates }`) so downstream
callers know to branch on `result.status` before reading byDepth/
summary. `context` is unchanged shape-wise (strict superset).
* New test: `context tool promotes top candidate via scoring when
multiple rows survive DB pre-filter`. The review flagged that the
existing file_path test works only because the mock ignores WHERE
parameters -- the scored-promotion path (top ≥ 0.95 AND gap > 0.09)
wasn't directly exercised. The new test uses two candidates both in
App.tsx-containing paths plus a kind hint so promotion is decided by
scoring, not DB pre-filtering. Also tightened the comment on the
earlier file_path test to describe the mock vs production divergence
honestly.
* NIT: added a paragraph explaining why `scored.length >= 2` is kept as
a defensive guard even though the `normalized.length === 1` early
return already covers the single-candidate path.
* Integration: two tests in `local-backend-calltool.test.ts` targeted
`'authenticate'`, which now correctly resolves as ambiguous (two
Method nodes: AuthService.authenticate and BaseService.authenticate).
Updated both to pass `file_path: 'src/auth.ts'` so they exercise the
new disambiguation API and still assert the METHOD_OVERRIDES filtering
they were originally about.
Edge case fix in the promotion gap check: IEEE754 makes 0.50 + 0.40 +
0.20 - 0.90 = 0.09999999999999998 instead of exactly 0.10, which would
otherwise break the "winner clearly dominates" intent for legitimate
1.00 vs 0.90 cases. Changed `>= 0.10` to `> 0.09`; same user-facing
intent, no floating-point sensitivity.
Verification (all from gitnexus/):
npx vitest run test/integration/class-impact-all-languages.test.ts
-> 52 pass (was 11 FAIL on CI before this fix)
npx vitest run test/integration/local-backend-calltool.test.ts
-> 18 pass (was 2 FAIL on CI before this fix)
npx vitest run test/integration/java-class-impact.test.ts
-> 10 pass (regression guard for #480 preserved)
npx vitest run test/unit/calltool-dispatch.test.ts
-> 65 pass (1 new test + 4 from original #470 PR)
npm run test:unit
-> 3626 pass, 4 pre-existing env failures unchanged
npx tsc --noEmit
-> clean
|
||
|
evolution
|
02739085d2
|
feat(embeddings): AST-aware chunking with offset-based splitting (#889)
Some checks are pending
CI / quality (push) Waiting to run
CI / tests (push) Waiting to run
CI / e2e (push) Waiting to run
CI / Save PR Metadata (push) Blocked by required conditions
CI / CI Gate (push) Blocked by required conditions
Release Candidate / Check if release candidate should run (push) Waiting to run
Release Candidate / ci (push) Blocked by required conditions
Release Candidate / Publish release candidate to npm (push) Blocked by required conditions
|
||
|
enih
|
eb0d9c51a0
|
fix: set env.cacheDir to user-writable location (#845)
When @huggingface/transformers is installed globally (e.g. via npm install -g), it defaults its cache directory to ./node_modules/.cache inside its own install dir, which is unwritable by non-root users. This causes EACCES errors on first use when the model is downloaded: EACCES: permission denied, mkdir '/usr/lib/node_modules/gitnexus/node_modules/@huggingface/transformers/.cache' Set env.cacheDir before pipeline() is called in both embedders (CLI and MCP). Respects HF_HOME env var if set, falls back to ~/.cache/huggingface. Co-authored-by: Sisyphus <sisyphus@gitnexus.dev> |
||
|
Mr. WorldwideBrown
|
4911201664
|
fix: map diff hunks to symbol line ranges in detect_changes (#779)
* fix: map diff hunks to symbol line ranges in detect_changes The detect_changes tool previously used `git diff --name-only` and picked the first 20 arbitrary symbols from each changed file. This produced false positives (unchanged symbols reported as modified) and false negatives (actually changed symbols dropped by the LIMIT). Now uses `git diff -U0` to get unified diff with hunk headers, parses the @@ line ranges, and queries for symbols whose [startLine, endLine] range overlaps the diff hunks. Only truly touched symbols are reported. Also fixed the CONTAINS path match to ENDS WITH to prevent cross-file false positives from substring matching. Fixes #758 * fix: address review feedback - variable shadowing, batch queries, tests - Rename `params` to `queryParams` in detectChanges hunk-mapping loop to avoid shadowing the outer method parameter - Replace N+1 per-symbol process lookup with a single batched query using WHERE n.id IN $ids (same pattern as impact BFS traversal) - Add unit tests for parseDiffHunks covering single/multi file, single/multi hunk, omitted count, pure-deletion, and empty input * style: fix prettier formatting in parse-diff-hunks test |
||
|
Gergő Magyar
|
0561d24efd
|
feat: METHOD_IMPLEMENTS edges, overload disambiguation, MethodExtractor unification (#574) (#642) | ||
|
Abhigyan Patwari
|
153262304c
|
fix(mcp): unify stdout silencing to prevent embedder/pool-adapter conflicts (#645) | ||
|
Gergő Magyar
|
63fc4c795f
|
feat: MethodExtractor configs for Python, PHP, Swift, Dart, Rust, Ruby (#624)
* feat: MethodExtractor configs for Python, PHP, Swift, Dart, Rust, Ruby with exhaustive integration tests Add per-language MethodExtractionConfig for all remaining tree-sitter languages (RFC #568 PR 2). Each config follows the established createMethodExtractor() factory pattern — no new types, no parse-worker changes. Configs: - Python: @abstractmethod, @staticmethod/@classmethod, *args/**kwargs, type hints, _/__ visibility - PHP: abstract/final/static keywords, PHP 8 #[] attributes, __construct/__destruct - Swift: 5-level visibility, protocol-as-abstract, static/class methods, @ attributes - Dart: _ convention visibility, abstract (no body), method_signature unwrapping - Rust: pub visibility, &self receiver, trait_item + impl_item, #[] attributes - Ruby: positional visibility via sibling-walk, singleton_method as static Integration fixtures (18 directories) covering 3 resolution patterns: - Method enrichment: parameterTypes, isAbstract, isFinal, annotations on graph nodes - Overload dispatch: arity-based CALLS resolution via parameterTypes - Abstract dispatch: abstract/concrete method distinction (Python, PHP, Rust, Swift) Go deferred — requires factory changes for receiver-based method extraction. Closes #571 * fix: address code review findings across 6 MethodExtractor configs Fix all actionable items from the PR #624 deep-dive review: Dart (critical — fixes 6 CI failures): - isDartStatic: check children first, siblings as fallback - isDartAbstract: handle declaration nodes for abstract methods - extractSingleParam: detect required keyword as sibling token - Add declaration to methodNodeTypes, mixin_declaration to typeDeclarationNodes - Add member call query for variable assignments in tree-sitter-queries Python: - hasDecorator now matches dotted paths (e.g. @abc.abstractmethod) - Fix version comment from ^0.23.6 to 0.23.4 PHP: - Add enum_declaration to typeDeclarationNodes (PHP 8.1+) - Add version comment for 0.23.12 Swift: - Add isOverride using hasKeyword/hasModifier pattern Rust: - Fix version comment from ^0.23.2 to 0.23.1 Also: identifier fallback in generic.ts for mixin owner names, Dart integration test label fix (Method vs Function), version comment for tree-sitter-dart 1.0.0. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: Dart extension_declaration and Ruby module_function support Dart: - Add extension_declaration to typeDeclarationNodes and extension_body to bodyNodeTypes — extension methods are now extracted into the graph - Add extension_declaration and mixin_declaration to CLASS_CONTAINER_TYPES for HAS_METHOD edge resolution Ruby: - module_function now maps to visibility 'private' in extractRubyVisibility - module_function methods marked isStatic via backward-walk in isStatic - Override semantics: private/public after module_function resets isStatic Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * feat(go): Go MethodExtractor config with receiver-based extraction Add Go as the 13th language with a per-language MethodExtractor config. Go methods are top-level (not nested in struct bodies), so this adds extractFromNode() to the MethodExtractor interface for direct method node extraction without an enclosing class. Config extracts: - Name from field_identifier (methods) / identifier (functions) - Return type including multi-return (first type from parameter_list) - Parameters with variadic support - Visibility via uppercase/lowercase convention - Receiver type with pointer unwrapping (*User → User) - isStatic for functions (no receiver) Infrastructure: - extractOwnerName optional hook on MethodExtractionConfig - extractFromNode on MethodExtractor (factory auto-implements) - Parse-worker uses extractFromNode when no enclosing class found - method_declaration added to CLASS_CONTAINER_TYPES Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * test: method enrichment integration tests for 7 languages + TS abstract class fix Add method-enrichment integration test fixtures and test blocks for Go, C++, Java, Kotlin, TypeScript, JavaScript, and C#. Each fixture tests: class detection, HAS_METHOD edges, EXTENDS edges, isAbstract, isStatic, annotations, parameterTypes, and CALLS edge resolution. Fixes found during testing: - Remove method_declaration from CLASS_CONTAINER_TYPES (added for Go but broke Java/C# HAS_METHOD edge resolution — method_declaration is also Java's method node type) - Add abstract_class_declaration query to TypeScript tree-sitter queries (was missing, so abstract classes were invisible to pipeline) 1699 integration tests pass across 20 test files, 0 regressions. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * style: format typeDeclarationNodes array for better readability in PHP config * fix: Go interface methods + Rust impl-for-Struct owner resolution Go: - Add method_elem to methodNodeTypes so interface method signatures are extractable as abstract methods - Integration test: Animal interface detected, Speak isAbstract, CALLS edges from app.go Rust: - Add extractOwnerName to resolve impl Trait for Struct to the concrete Struct (not the Trait) — fixes method misattribution - Fix findEnclosingClassId to generate Struct: label (not Impl:) for impl blocks so HAS_METHOD edges resolve to struct nodes - Tighten abstract-dispatch test: assert SqlRepo owns find/save generic.ts: - Fix extractOwnerName fallback: when hook returns a value, skip both name-field and type_identifier scan (was overwriting result) 1703 integration tests pass, 0 regressions. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: code review response — Rust impl label, Swift params, Dart async, sequential methodExtractor Address code review findings from PR #624: - ast-helpers: Rust `impl Trait for Struct` uses Struct label (matches existing graph node), plain `impl Struct` uses Impl label (matches definition.impl) - swift: fix parameter type extraction (user_type not type_annotation), detect default values as function_declaration siblings, add version comment - dart: isDartAsync now detects async*/sync* generators, add clarifying comment for declaration nodes in extension bodies - python: correct isFinal comment (PEP 591 @typing.final exists, just not modeled) - parsing-processor: port methodExtractor enrichment to sequential path so isAbstract/isStatic/visibility/annotations/isFinal populate on <15-file repos - tests: remove silent `if (prop !== undefined)` guards, assert properties directly, fix label queries (Dart Method vs Function, Swift Method for protocol methods), add Rust HAS_METHOD sourceLabel tests, Swift parameterTypes tests, and Dart async/sync* integration tests with fixture Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: Rust grammar gap + qualified method IDs to resolve same-file collisions Phase 1 — Rust grammar: - Add function_signature_item query to RUST_QUERIES so abstract trait methods (fn speak(&self) -> String;) become graph nodes with isAbstract=true Phase 2 — Qualified method IDs: - findEnclosingClassInfo returns {classId, className} for AST-based class lookup - Both parsing paths (sequential + worker) qualify method/property IDs with enclosing class: Method:file:ClassName.method instead of Method:file:method - extractFuncNameFromSourceId handles ClassName.method format - Fixes silent data loss when same-name methods in different classes shared a file (e.g., Animal.speak and Dog.speak both now exist as distinct graph nodes) Test updates: - Rust: abstract+concrete trait methods both verified, function count adjusted - Python: static method disambiguation now emits 2 CALLS edges (correct — no more ID collision masking the second call) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: owner-aware resolution for qualified method IDs Address Codex adversarial review findings after qualified ID change: - findEnclosingFunction: disambiguate candidates by ownerId when multiple same-name methods exist in file; qualify fallback-generated IDs - findEnclosingFunctionId (worker): qualify sourceIds with enclosing class name so CALLS source attribution matches definition-phase node IDs - buildExportedTypeMapFromGraph: use lookupExactAll + nodeId match instead of lookupExactFull which returns first definition for bare name Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> * fix: methodExtractor variadic arity, return type preservation, PHP abstract dispatch Three bugs in the methodExtractor enrichment path broke 17 integration tests: 1. Variadic parameterCount: buildMethodProps and parse-worker set parameterCount = info.parameters.length even for variadic functions, causing arity filtering to reject valid calls. Now checks isVariadic and sets parameterCount = undefined (matching extractMethodSignature). 2. C++ bare `...` token: extractCppParameters only iterated named children, missing the unnamed `...` token in C-style variadics like log_entry(const char* fmt, ...). Added fallback scan of all children. 3. Return type stripping: All 11 language extractReturnType functions used extractSimpleTypeName() which strips generic parameters (List<User> → "List", Task<User> → "Task"). Changed to .text?.trim() to preserve full generic types needed for for-loop iterable resolution, async-await binding, and return-type inference. Also fixes PHP abstract dispatch test that matched SqlRepository instead of the interface due to ambiguous filePath.includes('Repository') filter, and adds parent-walk fallback in PHP isAbstract for extractFromNode path. * chore: remove plan and review artifacts from PR * fix: address Round 4 review findings + infrastructure improvements - Ruby: add singleton_class support for class << self methods (4 new tests) - PHP: add enum_declaration to CLASS_CONTAINER_TYPES - Dart: add mixin/extension labels to CONTAINER_TYPE_TO_LABEL - Swift: add TODO for unverifiable struct/enum node types on Node 22 - C#: add grammar version comment (0.23.1) - Ruby: fix version comment range to pin (0.23.1) - Rust/ast-helpers: add cross-reference comments for impl_item duplication - ast-helpers: document CLASS_CONTAINER_TYPES ↔ typeDeclarationNodes invariant - generic.ts: replace Array.includes with Set for O(1) dedup in addNestedBodies - Go/Python/Ruby: align isAbstract signature with 2-param interface contract - CLAUDE.md: fix malformed backtick around gitnexus:start HTML comment - parsing-processor: add per-class method extraction cache (eliminates O(N*M)) - ast-helpers: add scoped_type_identifier to impl_item resolution - call-processor: add dev-mode warnings at silent candidates[0] fallbacks - MCP context(): surface methodMetadata for Method/Function/Constructor nodes - resources.ts: update schema to list all stored Method properties * fix: singleton_class HAS_METHOD edge regression in findEnclosingClassInfo singleton_class (class << self) was added to CLASS_CONTAINER_TYPES but has no name field — its receiver `self` has node type 'self', not 'identifier'. findEnclosingClassInfo now walks up to the enclosing class/module to inherit its name, matching ruby.ts:extractOwnerName. Also fixes findEnclosingClassNode in parse-worker.ts to skip singleton_class and return the actual class/module node. Adds integration test assertions for from_habitat (class << self method): HAS_METHOD edge from Animal, isStatic=true, parameterCount=1. --------- Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com> |
||
|
ivkond
|
4fed097abb |
feat(group): add sync pipeline, CLI, MCP tools, and monorepo fixture
Wire extractors into the sync pipeline with service boundary detection. GroupService provides high-level API for all group operations. - Sync pipeline: orchestrates extraction (HTTP, gRPC, topics) with service boundary assignment and exact matching - GroupService: groupList, groupSync, groupContracts, groupQuery, groupStatus (groupImpact deferred to cross-repo follow-up PR) - CLI: group create/add/remove/list/sync/contracts/query/status - MCP tools: group_list, group_sync, group_contracts, group_query, group_status - Monorepo fixture: 3 services (auth/orders/gateway) connected via gRPC + Kafka + HTTP — all intra-repo cross-links discovered - Documentation: CLI commands and MCP tools added to both READMEs Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> |
||
|
ivkond
|
52277247fe |
feat(group): add group infrastructure and contract matching
Core foundation for repository group analysis: - Type system: ContractType, ExtractedContract, StoredContract, CrossLink with optional `service` field for intra-repo matching - Config parser for group.yaml (repos, detection flags, matching thresholds) - Contract registry storage with atomic writes - Exact matching engine with per-type normalization (HTTP, gRPC, topic) and intra-repo support (different services within same repo can match) - Extract LadybugDB pool-adapter from MCP backend for reuse by sync pipeline - Git staleness checker for group status reporting Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com> |
||
|
Gergő Magyar
|
acf6fbdd39
|
feat: configure eslint with unused import removal (#564)
* feat: configure eslint with unused import removal Add ESLint v9 (flat config) for code quality: - eslint-plugin-unused-imports for auto-removing dead imports - @typescript-eslint for TypeScript-aware linting - eslint-plugin-react-hooks for React hooks rules - eslint-config-prettier to avoid formatting conflicts - lint-staged runs eslint --fix before prettier on .ts/.tsx - CI lint job added to ci-quality.yml * refactor: remove unused imports via eslint --fix Auto-fixed by eslint-plugin-unused-imports. No logic changes. * chore: add eslint fix commit to .git-blame-ignore-revs |
||
|
Gergő Magyar
|
bf09eab95b
|
feat: configure prettier with pre-commit hook (#563)
* feat: configure prettier with pre-commit hook integration Add prettier, lint-staged, and prettier-plugin-tailwindcss at the repo root with husky pre-commit hook integration. Moves husky from gitnexus/ to root package.json for reliable hook installation. - Root package.json with prepare/format/format:check scripts - .prettierrc with endOfLine:lf and tailwindStylesheet for TW v4 - .prettierignore excluding fixtures, vendor, generated, *.d.ts, *.md - .gitattributes enforcing LF line endings for Windows consistency - Pre-commit hook uses direct node_modules/.bin/ paths (no npx) * style: apply prettier formatting to entire codebase One-time bulk format. No logic changes. Use .git-blame-ignore-revs to skip this commit in git blame. * chore: add .git-blame-ignore-revs for prettier format commit * perf: pre-commit hook runs only tests related to staged files Use vitest --related to scope test execution to tests that import the changed files, instead of running the full suite on every commit. * perf: remove vitest from pre-commit hook, keep in CI only Pre-commit now runs lint-staged + tsc only. Tests run in CI (ci-tests.yml) where they belong — keeps commits fast. * ci: add prettier format check to quality workflow PRs will now fail if code isn't formatted with prettier. |
||
|
Gergő Magyar
|
fd7fb5bf1f
|
feat: unify web and cli ingestion pipeline (#536)
* feat: add server-side ingestion API (POST /api/analyze, SSE progress)
Extract core analysis orchestration from CLI into shared run-analyze.ts
module. Add server-side analyze endpoints so the web app can trigger
ingestion via HTTP instead of running the full pipeline in-browser.
New files:
- src/core/run-analyze.ts — shared runFullAnalysis() orchestrator
- src/server/analyze-job.ts — job manager (single-slot, dedup, SSE events)
- src/server/analyze-worker.ts — forked child process (8GB heap, IPC)
- src/server/git-clone.ts — shallow clone/pull with SSRF protection
API endpoints:
- POST /api/analyze — start analysis (returns 202 + jobId)
- GET /api/analyze/:jobId — poll job status
- GET /api/analyze/:jobId/progress — SSE progress stream
Security: URL validation blocks private IPs and non-HTTP schemes.
Path validation requires absolute paths. Git stderr not leaked to API.
* feat(web): add server-side analyze UI (Phase 2)
Add "Analyze on Server" flow to the web app's Server tab so users
can trigger server-side ingestion from the browser. On completion,
the graph is automatically loaded via the existing connectToServer flow.
New files:
- AnalyzeProgress.tsx — progress bar with phase label, elapsed time, cancel
Modified files:
- backend.ts — startAnalyze(), streamAnalyzeProgress() SSE client
- DropZone.tsx — analyze URL input + button below Connect section
- App.tsx — onServerAnalyze handler wires analyze -> connect flow
* feat: add job cancellation, timeout, and child process tracking (Phase 3)
- DELETE /api/analyze/:jobId — cancel running analysis (SIGTERM to worker)
- 30-minute timeout kills long-running workers automatically
- Child process refs tracked in JobManager for cleanup on shutdown
- dispose() kills all active children on SIGINT/SIGTERM
- Web cancel button now calls server DELETE endpoint
- cancelAnalyze() added to web backend client
* refactor(web): remove browser ingestion pipeline (Phase 4)
Delete 16 duplicated ingestion files, 2 unused service files
(git-clone, zip), and tree-sitter parser-loader from gitnexus-web.
All ingestion now runs server-side via POST /api/analyze.
Deleted (18 files, ~5,000 lines):
- core/ingestion/*.ts (16 pipeline processors)
- core/tree-sitter/parser-loader.ts (WASM tree-sitter loader)
- services/git-clone.ts (isomorphic-git client-side clone)
- services/zip.ts (JSZip extraction)
Simplified:
- DropZone.tsx — server-only (removed ZIP/GitHub tabs)
- ingestion.worker.ts — removed runPipeline/runPipelineFromFiles
- useAppState.tsx — removed pipeline callbacks
- App.tsx — removed handleFileSelect/handleGitClone
- main.tsx — removed Buffer polyfill for isomorphic-git
- types/pipeline.ts — removed PipelineResult/serialize helpers
Kept: cluster-enricher.ts (LLM enrichment, still used by worker)
Dependencies now removable: web-tree-sitter, isomorphic-git,
@isomorphic-git/lightning-fs, jszip (estimated 3-4MB bundle savings)
* refactor(web): sync graph schema from CLI + delete WASM grammars
Sync graph/types.ts and lbug/schema.ts from the CLI (source of truth)
to the web module so the browser LadybugDB can handle all node and
relationship types the server pipeline produces.
Synced types: Route, Tool, Section node labels; HANDLES_ROUTE, FETCHES,
HANDLES_TOOL, ENTRY_POINT_OF, WRAPS, QUERIES relationship types;
description fields on Function/Class/Interface/Method/CodeElement.
Deleted: public/wasm/ directory (14 tree-sitter WASM grammars + core).
Removed deps: web-tree-sitter, isomorphic-git, @isomorphic-git/lightning-fs,
jszip, buffer, @types/jszip (~3-4MB bundle savings).
* feat: create gitnexus-shared package for unified type definitions
Create a new gitnexus-shared package that is the single source of truth
for types shared between the CLI and web modules:
- SupportedLanguages enum (15 languages)
- Graph types: NodeLabel, NodeProperties, RelationshipType, GraphNode, GraphRelationship
- Schema constants: NODE_TABLES, REL_TYPES, REL_TABLE_NAME, EMBEDDING_TABLE_NAME
- Pipeline types: PipelinePhase, PipelineProgress
Both gitnexus (CLI) and gitnexus-web import from gitnexus-shared via
file: dependency. Each package re-exports and extends with platform-specific
additions (CLI: KnowledgeGraph with mutation methods; Web: simpler KnowledgeGraph).
This ensures types can never drift between packages — adding a new
language, node type, or relationship type in gitnexus-shared automatically
propagates to both consumers.
* refactor: import shared types directly from gitnexus-shared at call sites
Replace all re-export patterns with direct imports from gitnexus-shared.
72 files updated across CLI and web:
- SupportedLanguages: 49 CLI files now import from 'gitnexus-shared'
instead of '../config/supported-languages.js'
- GraphNode, GraphRelationship, NodeLabel: 22 CLI + 10 web files now
import from 'gitnexus-shared' instead of local re-export wrappers
- NODE_TABLES: api.ts imports from 'gitnexus-shared'
- PipelineProgress: useAppState.tsx imports from 'gitnexus-shared'
Local types.ts files now only define platform-specific KnowledgeGraph
(CLI has mutation methods, web has add-only). No more re-exports.
* fix: update lock files for gitnexus-shared, remove stale vite polyfills
Add gitnexus-shared@1.0.0 to lock files so npm ci succeeds in CI.
Remove buffer polyfill and global define from vite.config.ts (isomorphic-git was removed).
* fix(security): add write guard to HTTP /api/query, fix CORS proxy bypass
- Add isWriteQuery() check to POST /api/query handler — blocks CREATE,
DELETE, SET, MERGE, DROP, etc. via HTTP API (guard was only in MCP
pool adapter and browser-side, not the HTTP server path)
- Extend CYPHER_WRITE_RE with CALL, INSTALL, LOAD keywords
- Fix CORS proxy subdomain bypass: endsWith('github.com') allowed
'evil-github.com'. Now requires exact match or '.github.com' suffix
* feat(server): enhance /api/search with enrichment, add /api/grep, strip graph content
- POST /api/search: add mode param (hybrid|semantic|bm25), server-side
enrichment returns connections/cluster/processes per result in one call
(collapses 31 sequential HTTP calls to 1 for the agent search tool)
- GET /api/grep: regex search across indexed file contents, eliminates
need to transfer all file contents to browser
- GET /api/graph: strip content field by default (80-95% payload
reduction). Use ?includeContent=true for backward compat
- Add LRU cache invalidation hook point for future caching
* feat(server): add /api/embed endpoint for server-side embedding generation
- POST /api/embed: triggers embedding pipeline via onnxruntime-node
with JobManager for single-slot concurrency, timeout, and dedup
- GET /api/embed/:jobId: poll job status
- GET /api/embed/:jobId/progress: SSE stream with heartbeat, event IDs,
and X-Accel-Buffering:no header for proxy compatibility
- DELETE /api/embed/:jobId: cancel running embedding job
- Maps embedding pipeline phases (ready→complete, error→failed) to
JobManager status conventions
* feat(web): create consolidated BackendClient module
Single HTTP client replacing backend.ts, server-connection.ts, and
worker HTTP helpers. Includes:
- Typed methods: runQuery, search (enriched), grep, readFile, connect
- Generic streamSSE<T> utility extracted from analyze progress pattern
- BackendError with discriminated code field (network/server/client/timeout)
- Embed API: startEmbeddings, streamEmbeddingProgress, cancelEmbeddings
- Search with mode param (hybrid|semantic|bm25) and enrichment
* refactor(web): rewrite Graph RAG tools for backend-only HTTP queries
- Search tool: uses enriched /api/search (1 call replaces 31 sequential queries)
- Cypher tool: removes browser-side embedding; {{QUERY_VECTOR}} routes to
/api/search with mode:'semantic' instead of local transformers.js
- Grep tool: uses /api/grep instead of in-memory fileContents map
- Read tool: uses /api/file instead of fileContents map lookup
- Impact tool: getCallSiteSnippet now async via /api/file
- createGraphRAGTools now accepts GraphRAGBackend interface instead of
7 separate function params + fileContents map
- createGraphRAGAgent simplified to (config, backend, context?)
- Removed imports: embedder, lbug/schema (replaced with gitnexus-shared)
- Net: -205 lines
* refactor(web): delete WASM infrastructure, remove 7 packages (-5242 lines)
Delete browser-side LadybugDB, embeddings, search, and worker:
- gitnexus-web/src/core/lbug/ (adapter, csv-generator, schema, query-result)
- gitnexus-web/src/core/embeddings/ (embedder, pipeline, text-gen, types)
- gitnexus-web/src/core/search/ (bm25-index, hybrid-search)
- gitnexus-web/src/workers/ingestion.worker.ts (828 lines)
- gitnexus-web/src/services/server-connection.ts (merged into backend-client)
- gitnexus-web/src/types/lbug-wasm.d.ts
Remove packages: @ladybugdb/wasm-core, @huggingface/transformers,
comlink, minisearch, vite-plugin-wasm, vite-plugin-top-level-await,
vite-plugin-static-copy
Update vite.config.ts: remove WASM plugins, COOP/COEP headers,
worker config, optimizeDeps exclude
Update imports: App.tsx, DropZone, Header, AnalyzeProgress,
BackendRepoSelector, useBackend → backend-client
* refactor(web): replace Worker/Comlink with direct BackendClient calls
- useAppState: remove Worker instantiation, Comlink.wrap, apiRef.
All queries now go through BackendClient HTTP functions directly.
- Agent runs on main thread (I/O-bound LLM streaming, not CPU-bound)
- initializeAgent: creates GraphRAGAgent with GraphRAGBackend interface
bound to BackendClient methods (runQuery, search, grep, readFile)
- startEmbeddings: calls POST /api/embed + SSE progress instead of
running browser-side transformers.js pipeline
- switchRepo: no longer loads graph into WASM DB or extracts fileContents
- App.tsx: handleServerConnect simplified (no fileContents, no loadServerGraph)
- Delete old backend.ts (replaced by backend-client.ts)
- Net: -396 lines
* fix(web): fix await-in-map build error in agent streaming
Move dynamic import of AIMessage outside .map() callback to avoid
"await can only be used inside an async function" build error.
* fix(web): remove stale apiRef references that broke chat functionality
sendChatMessage referenced apiRef.current (deleted Worker ref) which
would throw TypeError. Replaced with agentRef.current guard since agent
now runs on main thread.
* fix(server): dispose embedJobManager on shutdown, fix job mutation
- Add embedJobManager.dispose() to shutdown handler (was missing,
causing cleanup timer to keep Node process alive)
- Replace direct job.repoName/status mutation with updateJob() to
ensure SSE event emission for initial status change
* fix(server): parameterize Cypher, harden grep, unify SSE endpoints
- Search enrichment: replace string interpolation with executePrepared()
using $nid parameter binding to prevent Cypher injection
- Add executePrepared() to core lbug-adapter (prepare/execute pattern)
- /api/grep: add 200-char pattern length limit (ReDoS protection),
search files on disk instead of loading entire corpus into memory
(constant memory usage regardless of repo size)
- Extract mountSSEProgress() shared helper for SSE streaming — both
analyze and embed endpoints now have consistent heartbeat (30s),
event IDs (reconnection support), and X-Accel-Buffering header
* refactor(web): remove dead code from Worker-era architecture
- Remove loadServerGraph no-op function, interface member, and all consumers
- Remove testArrayParams stub and interface member
- Remove fileContents state from GraphStateProvider (never populated in
server-side architecture)
- Remove forceDevice parameter from startEmbeddings (server-side, no device choice)
- Replace phantom EmbeddingProgress type with inline { phase, percent }
- Replace resolvePathFromContents (needed fileContents Map) with graph-based
file path resolution using filePathIndex built from graph nodes
- Fix: AI citation grounding ([[file.ts:10]]) now works via graph node lookup
instead of broken fileContents-based resolution
* fix(web): use streamAgentResponse for full tool_call/reasoning streaming
Replace naive agent.stream() loop that only handled content chunks with
streamAgentResponse() generator from agent.ts. This properly routes:
- reasoning tokens (before/between tool calls)
- tool_call events (name, args, status)
- tool_result events (completed tool output)
- content tokens (final answer after all tools done)
Previously the onChunk handler for tool_call/tool_result/reasoning was
dead code since the streaming loop only emitted content events.
* fix(web): resolve CI type errors from dead code removal
- Import GraphNode/GraphRelationship from gitnexus-shared in graph.ts
(not re-exported from local types.ts)
- Add Route, Tool entries to NODE_COLORS and NODE_SIZES constants
- Add PipelineResult type to web types/pipeline.ts
- Remove fileContents from CodeReferencesPanel and RightPanel
- Remove testArrayParams and forceDevice from EmbeddingStatus
- Remove forceDevice args from startEmbeddings() calls in App.tsx
- Fix embeddingProgress property accesses for simplified type
* fix(ci): add setup-gitnexus-web action, build shared once per job
- Remove prepare script from gitnexus-shared (tsc not available during
npm ci of consuming packages)
- Create .github/actions/setup-gitnexus-web composite action: builds
gitnexus-shared then runs npm ci for gitnexus-web
- setup-gitnexus action: already builds gitnexus-shared for CLI jobs
- ci-quality typecheck-web: uses setup-gitnexus-web (DRY)
- ci-e2e: uses setup-gitnexus-web (DRY)
- ci-tests: gitnexus-shared already built by setup-gitnexus, just
install web deps without rebuilding
* fix(ci): use prepare script so gitnexus-shared builds during npm ci
Move typescript from devDependencies to dependencies in gitnexus-shared
so the prepare script (tsc) works when npm resolves file: deps during
npm ci. No GHA modifications needed — npm handles the build lifecycle
automatically.
Remove manual gitnexus-shared build steps from setup-gitnexus and
setup-gitnexus-web actions.
* fix(ci): build gitnexus-shared explicitly in setup actions
The file: dependency protocol doesn't reliably run prepare scripts
because devDependencies aren't installed first. Instead of fragile
lifecycle hacks, build gitnexus-shared explicitly in both setup actions:
- setup-gitnexus: npm install && npm run build in gitnexus-shared/
- setup-gitnexus-web: same, before npm ci in gitnexus-web/
- ci-tests: shared already built by setup-gitnexus, web just npm ci
No prepare script, no dist in git, no typescript as a prod dependency.
* fix: remove CALL from CYPHER_WRITE_RE — breaks FTS and vector search
CALL is used by read-only procedures: CALL QUERY_FTS_INDEX(...) and
CALL QUERY_VECTOR_INDEX(...). Adding it to the write guard blocked all
FTS search, causing 3 test failures. The database is opened in read-only
mode as defense-in-depth against write procedures via CALL.
Keep INSTALL and LOAD in the blocklist (genuinely dangerous).
* fix(web): update vercel.json for gitnexus-shared, remove COOP/COEP
- Add installCommand that builds gitnexus-shared before installing
web deps (Vercel doesn't know about the monorepo file: dependency)
- Remove Cross-Origin-Opener-Policy and Cross-Origin-Embedder-Policy
headers (no longer needed — WASM LadybugDB removed)
* fix(web): update tests for deleted modules
- Delete csv-generator.test.ts (tests deleted WASM-only csv-generator)
- Update security-guards.test.ts: import NODE_TABLES/REL_TYPES from
gitnexus-shared instead of deleted src/core/lbug/schema
- Update server-connection.test.ts: import normalizeServerUrl from
backend-client, remove extractFileContents tests (function deleted)
* fix(e2e): remove Server tab click — UI is now server-only
The DropZone no longer has ZIP/GitHub/Server tabs (browser ingestion
was removed). The server URL input is directly visible on the landing
page. Update e2e test to skip the tab click and go straight to input.
All 5 e2e tests pass locally.
* refactor: use gitnexus-shared for PipelinePhase/PipelineProgress types
CLI was duplicating PipelinePhase and PipelineProgress locally instead
of importing from gitnexus-shared. Updated all consumers to import
directly. Also removed dead code: SerializablePipelineResult,
serializePipelineResult(), deserializePipelineResult().
* fix(server): address PR #536 review — security, race conditions, dead code
- Fix path traversal in POST /api/analyze: split into isAbsolute + normalize check
- Add shared repo lock (activeRepoPaths) preventing concurrent analyze+embed on same repo
- Fix 202 response returning actual job.status instead of hardcoded 'queued'
- Add 30-minute timeout for embedding jobs (was missing unlike analyze jobs)
- Fix DropZone calling startAnalyze without setting backend URL first
- Add SSE reconnect with exponential backoff (3 retries) and Last-Event-ID
- Fix normalizeServerUrl to return base URL (no /api suffix) — clear contract
- Delete dead code: proxy.ts, server-graph-hydration.ts, pipeline.ts re-export barrel
- Update LoadingOverlay to import PipelineProgress directly from gitnexus-shared
* fix(server): fix repo lock key mismatch and embed cancel race
- Use getStoragePath(targetPath) as lock key in analyze handler to match
embed handler's entry.storagePath — keys now always align
- Guard embed completion: don't overwrite 'failed' with 'complete' when
job was cancelled while pipeline was still running
- Remove unused jobType parameter from acquireRepoLock
- Log backend.init() errors instead of silently swallowing
* fix: add gitnexus-shared as a local dependency in package-lock.json
* refactor: move language detection to gitnexus-shared, add syntax highlighting for all 15 languages
Move getLanguageFromFilename() from CLI to gitnexus-shared with COBOL
support added. Add getSyntaxLanguageFromFilename() for Prism-compatible
syntax highlighting covering all 15 code languages plus auxiliary
formats (json, yaml, markdown, html, css, bash, sql, xml).
Refactor CodeReferencesPanel to use shared function instead of a local
30-line switch. Delete dead gitnexus-web/src/config/supported-languages.ts
(web already imports SupportedLanguages from gitnexus-shared).
* feat(web): add first-time user onboarding with auto server detection
Replace the manual "Connect to Server" panel with an automatic onboarding
flow that guides first-time users through starting the GitNexus server.
Server detection:
- useBackend hook polls via setTimeout chain (3s, no overlap)
- Page Visibility API pauses polling when tab is hidden
- SSE heartbeat (/api/heartbeat) for instant disconnect detection
Onboarding UI (OnboardingGuide.tsx):
- Step-by-step flow: copy command → run → auto-connect
- Smart command: shows `gitnexus serve` in dev, `npx gitnexus@latest serve` in prod
- Node.js version auto-detected from package.json via Vite define
- Faux terminal windows with copy-to-clipboard, platform tabs, polling indicator
Transitions (DropZone.tsx):
- Crossfade wrapper with snapshot pattern for smooth phase transitions
- Three phases: onboarding → success (1.2s hold) → loading → graph
- Auto-recovery: falls back to onboarding if server dies or connect fails
Server changes:
- GET /api/heartbeat: SSE endpoint for liveness detection
- GET /api/info: version, launch context, Node.js version
- npm run serve script for local development
- app.disable('x-powered-by') hardening
* feat(web): add repo analysis UI, SSE heartbeat, and review fixes
Repo analysis:
- AnalyzeOnboarding: empty-state card when server has zero repos
- RepoAnalyzer: GitHub URL + Local Folder tabs with browse button
- Header repo dropdown: click project badge to switch repos or analyze new
- DropZone 'analyze' phase integrated into Crossfade transitions
Reliability fixes from 5-agent review:
- Polling: stop scheduling timers when tab hidden, restart on visibility return
- Heartbeat: exponential backoff (1s/2s/4s, 3 retries) prevents graph loss on blip
- RepoAnalyzer: completion timer tracked in ref, cleaned up on unmount
- DropZone: standardized card padding (p-7), heading sizes (text-lg)
Accessibility:
- prefers-reduced-motion global CSS rule (WCAG 2.3.3)
- focus-visible rings on CopyButton
- cursor-pointer on all Header buttons
- Consistent rounded-xl on all dropdowns
Cleanup:
- Deleted dead AnalyzeSheet.tsx (219 LOC) and BackendRepoSelector.tsx (89 LOC)
- Fixed AnalyzeProgress lucide import (lucide-react → @/lib/lucide-icons)
* fix(server): resolve analyze worker fork crash in dev mode
The forked analyze worker was crashing immediately with exit code 1
when running via `npm run serve` (tsx). Two issues:
1. Worker path resolved to `analyze-worker.js` but only `.ts` exists
in the source directory — the `.js` file is only in `dist/`.
2. On Windows, bare `--import tsx` in execArgv fails because Node's
ESM resolver for --import uses the child's CWD, not the parent's
node_modules. Windows also rejects raw paths as `d:` is not a
valid URL scheme.
Fix: detect dev vs prod via `import.meta.url` extension. In dev mode,
resolve `tsx/esm` to an absolute `file://` URL via `pathToFileURL()`
anchored to the parent's `createRequire` context. This works on all
platforms and doesn't depend on the child's CWD or PATH.
Also captures child stderr for better crash diagnostics.
Verified: `POST /api/analyze` with GitHub URL completes successfully
in dev mode (tsx) — status goes from cloning → analyzing → complete.
* fix(server): add worker auto-retry, error handling, and crash diagnostics
Worker resilience:
- Auto-retry up to 2 times with exponential backoff (1s, 2s) on crash
- SSE progress shows "Retrying after crash (1/2)..." during retry
- Captures child stderr for crash diagnostics in failure message
- AnalyzeJob tracks retryCount per job
Server error handling:
- app.listen wrapped in Promise so EADDRINUSE/EACCES propagate cleanly
- serve.ts catches startup errors with friendly messages and exit code 1
- EADDRINUSE gets actionable guidance (stop other process or --port flag)
- Global uncaughtException/unhandledRejection handlers prevent silent exits
- DEBUG=1 env var shows full stack traces
* feat: add e2e tests for onboarding flows, worker retry, and error handling
E2E tests (onboarding.spec.ts — 11 tests):
- Flow 1: OnboardingGuide shown when server unreachable (6 tests)
- Flow 2: Auto-connect with success card, analyze phase for zero repos
- Flow 3: Analyze form — GitHub URL validation, Local Folder tab, tab switching
- Flow 4: Repo dropdown in exploring view (skipped without live server)
Updated server-connect.spec.ts:
- Replaced manual Connect button flow with auto-connect waitForGraphLoaded
Server resilience:
- Worker auto-retry (2 attempts with exponential backoff) on crash
- Friendly error messages for serve startup failures (EADDRINUSE etc.)
- Global uncaughtException/unhandledRejection handlers prevent silent exits
- app.listen wrapped in Promise for proper error propagation
* refactor(shared): enforce exhaustive language coverage via Record types
Replace the if/else chain in getLanguageFromFilename with two exhaustive
Record<SupportedLanguages, ...> maps:
- EXTENSION_MAP: every language → its file extensions
- SYNTAX_MAP: every language → its Prism syntax identifier
Adding a new member to the SupportedLanguages enum without adding it to
both maps now produces a TypeScript compile error:
Property '[SupportedLanguages.NewLang]' is missing in type...
This matches the existing pattern in languages/index.ts (providers table)
which already uses `satisfies Record<SupportedLanguages, LanguageProvider>`.
Three compile-time enforcement points now exist:
1. EXTENSION_MAP in language-detection.ts (file extensions)
2. SYNTAX_MAP in language-detection.ts (Prism syntax identifiers)
3. providers in languages/index.ts (LanguageProvider instances)
* feat(web): load source code from server and scroll to selected line
CodeReferencesPanel now fetches file content via GET /api/file when a
node is selected, instead of showing "Code not available in memory".
- Fetches via readFile() from backend-client when selectedFilePath changes
- Shows loading spinner while fetching
- After content loads, auto-scrolls to the selected node's startLine
- Highlights the selected line range with a cyan left border
- Cancels in-flight fetch if selection changes before it completes
Also: refactored language-detection.ts to use exhaustive Record types
(EXTENSION_MAP and SYNTAX_MAP) so adding a new SupportedLanguages enum
member without implementing extensions/syntax is a compile error.
* feat: buffered file reading for Code Inspector
Server: GET /api/file now supports ?startLine=N&endLine=M for reading
a line range instead of the entire file. Returns { content, startLine,
endLine, totalLines }.
Client: readFile() returns ReadFileResult with metadata. When selecting
a symbol (function, class, method), fetches only ±50 lines around the
symbol's startLine/endLine instead of the full file. File nodes still
fetch the entire file.
SyntaxHighlighter startingLineNumber set from the buffer offset so line
numbers are correct even for partial reads.
* fix: adapt readFile callers to new ReadFileResult return type
tools.ts: readFile comes from GraphRAGBackend interface which returns
Promise<string> (the adapter in useAppState extracts .content), so
revert the { content } destructuring back to plain string assignment.
useAppState.tsx: wrap backendReadFile with { repo } options object
and extract .content to satisfy the GraphRAGBackend interface.
* fix(web): ensure new repos appear in list immediately after analysis
Two fixes:
1. DropZone: handleAnalyzeComplete now passes the repoName through to
connectToServer so the specific newly-analyzed repo loads — not the
server's default first repo.
2. App.tsx: fetchRepos() is now awaited BEFORE handleServerConnect in
both the DropZone and Header flows. This ensures the repo list is
populated before the exploring view renders, so the new repo appears
in the header dropdown immediately without a page reload.
* feat: delete repos, re-analyze with force, select after analysis
Server — DELETE /api/repo:
- Acquires repo lock first (409 if analyze/embed in flight)
- Closes LadybugDB, deletes index + clone dir, unregisters, re-inits
- Lock released in finally block
Server — analyze complete:
- backend.init() must succeed before SSE complete fires
- If backend.init() fails, job is marked failed (not complete)
Web — Header repo dropdown:
- Re-analyze: calls POST /api/analyze with force=true, shows spinning
icon + inline progress bar via SSE
- Delete: acquires lock, aborts any running re-analysis SSE for same
repo, refreshes list, switches to next repo
- After analysis completes: refreshes repo list, connects to the
specific repo by name, loads graph, shows in explorer
- Retry with 1.5s backoff on 404 (server may still be reinitializing)
Type safety:
- err: any → err: unknown + instanceof BackendError in retry loop
- Added missing BackendRepo + BackendError imports in App.tsx
|
||
|
ThinhKVT
|
aec9a3f216
|
[CLI] Fixes a false-positive in the Cypher write-detection regex and improves the Impact tool's enrichment path by using batched chunking and entry-point grouping (#496) (#507) | ||
|
marxo126
|
5a7ac218df
|
fix: shape_check false positives — quoted keys, DOM leaks, errorKeys (#501) | ||
|
chirag-nighut
|
fca7815c16 |
feat(wiki): add Cursor CLI as LLM provider option
Add Cursor headless CLI as a 4th provider option for wiki generation, allowing users to leverage their Cursor subscription for wiki pages. - Add --provider cursor flag and cursor-client.ts - Add --review flag for interactive module tree editing - Add --verbose flag for debugging - Improve module tree generation (flatten single-child, unique slugs) - Prevent DB timeout during long LLM calls Usage: gitnexus wiki --provider cursor --model claude-4.5-opus-high Made-with: Cursor |
||
|
Yash
|
a191c26571
|
fix(#480): resolve impact/context returning empty results for Java cl… (#489) | ||
|
marxo126
|
c437acf6bb
|
feat: deep flow detection — consumer access tracking, middleware chains, error shapes, api_impact tool (#482) | ||
|
Zander Raycraft
|
a3fac2f672
|
Merge pull request #395 from zm2231/feat/http-embedding-backend | ||
|
Gergő Magyar
|
907440cf0b
|
Merge pull request #427 from ShunsukeHayashi/fix/impact-confidence-412 | ||
|
Shunsuke Hayashi
|
d206bf6772 |
fix(ingestion): calculate confidence per resolution tier for heritage/MRO edges (#412)
Instead of hardcoding confidence: 1.0, compute it at ingestion time using
the same resolution tier system that CALLS edges already use.
Heritage edges (EXTENDS, IMPLEMENTS):
- resolveHeritageId now returns { id, confidence } using TIER_CONFIDENCE
- Same-file → 0.95, import-scoped → 0.9, global → 0.5
- Edge confidence = geometric mean of source and target confidence
(principled for partially-correlated cross-scope estimates, per
Dillig et al. POPL 2011 and Dempster-Shafer theory)
MRO edges (OVERRIDES):
- MRO-ordered → 0.9, class method wins → 0.95
- Single interface → 0.85, ambiguous/unresolved → 0.5
IMPORTS and CONTAINS intentionally keep 1.0 (deterministic).
Closes #412
|
||
|
Shunsuke Hayashi
|
b272c6864c |
docs(schema): add Community and Process node properties to cypher tool description (#411)
The cypher tool description and schema resource omit Community and Process node properties, causing agents to write failing queries on first attempt. Added property listings sourced from the actual LadybugDB schema definitions: - Community: heuristicLabel, cohesion, symbolCount, keywords, description, enrichedBy - Process: heuristicLabel, processType, stepCount, communities, entryPointId, terminalId Closes #411 |
||
|
zm2231
|
c8480f899d |
fix: address review feedback — deduplicate HTTP client, remove config cache, add guards
- Extract shared HTTP client (http-client.ts) used by both core and MCP embedders - Remove module-level httpConfig cache — read env vars fresh on every call so config set after module load (e.g. via dotenv) takes effect - Add NaN/non-positive guard on GITNEXUS_EMBEDDING_DIMS in schema.ts - Include scrubbed URL and batch index in error messages (no API key) - Wrap fetch rejections (DNS/timeout/connection) with same scrubbed context - MCP embedder delegates to shared httpEmbedQuery() instead of inline logic - apiKey confined to http-client.ts internals — not exported in any type or accessor - Remove HttpEmbeddingConfig from types.ts (replaced by internal HttpConfig) - All 16 HTTP embedder tests pass, tsc clean |
||
|
hiromima
|
893f77ae89 |
fix: address review feedback — watchdog, TOCTOU race, null guard, platform guard
- Watchdog timer now exempts in-flight queries via activeQueryCount, preventing premature stdout restoration during long queries (>1s) - Stale detection uses reinitPromises Map to prevent TOCTOU race where concurrent callers double-close the connection pool - Throttle meta.json staleness checks to once per 5s per repo - Add null guard for i.id in IN-clause construction - Enrichment queries run in parallel on non-arm64 platforms to preserve performance; sequential only on arm64 macOS where SIGSEGV occurs |
||
|
zm2231
|
7dcafb647b |
fix: address review feedback — timeout, retry, guards, tests
- Add AbortSignal.timeout(30s) on all fetch calls - Add retry with backoff for 429/5xx (core: 2 retries, MCP: 1 retry) - Guard initEmbedder() and getEmbedder() to throw in HTTP mode - Discard cached embeddings on dimension mismatch during incremental re-index - Add MCP embedQuery retry for transient failures - Add 16 unit tests covering both core and MCP HTTP paths - Fix README: concise, accurate env var docs |
||
|
zm2231
|
5dcc567870 |
fix: address code review — timeout, retry, safe initEmbedder guard
- Add AbortSignal.timeout(30s) to all HTTP fetch calls - Add retry with backoff for 429/5xx (max 2 retries) - initEmbedder() throws descriptive error in HTTP mode instead of null cast - embedding-pipeline.ts guards initEmbedder() behind isEmbedderReady() check - Error messages omit request details to avoid leaking API keys in logs |
||
|
hiromima
|
999fbf5b11 |
fix: sequential enrichment queries + stale data detection
Fixes three related issues that cause SIGSEGV crashes and stale data: 1. Impact enrichment queries (Promise.all → sequential await) The impact() method ran 3 enrichment queries concurrently via Promise.all against the same LadybugDB connection pool. On arm64 macOS, concurrent native DB access triggers SIGSEGV. Changed to sequential await. Also caps IN-clause to 100 IDs to prevent oversized queries. (#285, #290, #292) 2. Silence stdout during query execution silenceStdout()/restoreStdout() only wrapped createConnection() and initLbug(). Now also wraps executeQuery() and executeParameterized() to prevent native stdout writes from corrupting the MCP stdio stream during all DB operations. (#285) 3. Stale data after re-index ensureInitialized() checked pool existence but never verified whether the underlying index was rebuilt. Now reads meta.json's indexedAt timestamp on each call and closes/re-opens the pool when the index has changed. (#297) |
||
|
zm2231
|
c79c717790 |
fix: dynamic embedding dimensions for HTTP backend
- schema.ts: FLOAT[${EMBEDDING_DIMS}] reads from GITNEXUS_EMBEDDING_DIMS env
- embedding-pipeline.ts: vector search CAST uses actual query vector length
- mcp/core/embedder.ts: HTTP embedding support for MCP query-time search
Without this, using a 1024d model (e.g. bge-large) fails with
'Expected: 384, Actual: 1024' on LadybugDB vector insert.
|
||
|
Gergo Magyar
|
7b71b64427 |
fix(mcp): update tool descriptions for Phase 9C capabilities
- context tool: remove outdated "Phase 2" ACCESSES reference, document that CALLS edges resolve through field/method chains - cypher tool: fix Property query example to use declaredType (not description) - schema resource: add node_properties section documenting Method returnType, Property declaredType, Function parameterCount etc. - schema resource: clarify ACCESSES edge read/write coverage |
||
|
Abhigyan Patwari
|
60c93d7d4a
|
feat: upgrade @ladybugdb/core to 0.15.2 and remove segfault workarounds (#374)
* feat: upgrade @ladybugdb/core to 0.15.2 and remove segfault workarounds
The upstream fix (ladybug-nodejs#1) resolves the child QueryResult lifetime
segfault, making .close() safe on all platforms. This removes 6 workaround
sites:
- Remove `dangerouslyIgnoreUnhandledErrors` from vitest config
- Remove platform-conditional .close() guards in global-setup and test helper
- Delete test/setup.ts (process._getActiveHandles unref hack)
- Replace no-op cleanup in test-indexed-db.ts with real adapter close
- Fix pool adapter closeOne() to properly close connections with shared
Database refcount guard and orphaned connection handling in checkin()
- Update segfault-related comments across the codebase
Also bumps @ladybugdb/wasm-core to ^0.15.2 in gitnexus-web for consistency.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* fix: keep dangerouslyIgnoreUnhandledErrors for macOS N-API exit crash
The N-API destructor ordering crash during worker fork exit on macOS is
independent of the QueryResult lifetime fix in 0.15.2. Tests pass, but
the exit triggers a crash. Keep the flag with an updated comment
explaining the actual cause. Can be removed once LadybugDB fixes all
destructor ordering issues upstream.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
* ci: unify test run for single-pass coverage
- Update `npm test` to run all tests (unit + integration + lbug-db)
via `vitest run` instead of `vitest run test/unit`
- Add `test:unit` script for running unit tests only
- Remove `ci-integration.yml` — the per-file lbug-db process isolation
is no longer needed with `dangerouslyIgnoreUnhandledErrors` and
`fileParallelism: false` handling fork exit issues
- Update `ci-unit-tests.yml` to run all tests with build + coverage
- Simplify `ci.yml` gate (two jobs: quality + tests)
- Simplify `ci-report.yml` (single coverage artifact, no merge step)
* fix: update cli-commands test for renamed test:all → test:unit script
* fix: set USERPROFILE in setup-skills test for Windows compatibility
os.homedir() checks USERPROFILE on Windows, not HOME.
* fix: add isolate: false to lbug-db project to prevent fork crashes
On macOS, N-API destructors crash fork workers on exit. With
isolate: true (default), vitest recycles the fork between files,
triggering the crash after each file. After several crashes, the
remaining lbug-db files never execute.
isolate: false keeps all 8 lbug-db files in a single fork — the
fork only exits once after all files complete, and that single exit
crash is caught by dangerouslyIgnoreUnhandledErrors.
* fix: add unique sequence.groupOrder to vitest projects
Vitest v4 requires unique groupOrder when projects have different
maxWorkers (lbug-db has fileParallelism: false → maxWorkers: 1).
* fix: await async close() in global-setup and remove isolate: false
global-setup.ts called conn.close() and db.close() without await —
these return Promise<void> in @ladybugdb/core 0.15.2. The setup
function returned before the DB was fully closed, so vitest forks
hit a stale file lock when opening the same DB path, crashing the
lbug-db worker before any test ran.
isolate: false caused native state corruption after 2-3 open/close
cycles in the same fork (vitest-specific, not reproducible in plain
Node.js). Without it, each file gets its own module scope and the
N-API destructor crash at fork exit is caught by
dangerouslyIgnoreUnhandledErrors.
Also fixes fire-and-forget close() calls in the pool adapter —
try/catch around an async close() never catches rejections; changed
to .catch(() => {}) for proper unhandled-rejection prevention.
Before: 0/8 lbug-db files ran on macOS CI (fork crash).
After: 8/8 pass, 84 files, 3077 tests, zero errors.
* fix: update project index references in AGENTS.md and CLAUDE.md to reflect correct symbol counts and relationships
* feat: enhance lbug adapter with external database support and write operation validation
* feat: create ci-tests workflow for comprehensive test coverage across platforms
* ci: move PR report inline to ci.yml, delete ci-report.yml
The old ci-report.yml used workflow_run which always runs code from
the default branch (main). This meant the PR comment used main's
stale report template that still referenced the old unit/integration
split architecture — causing "Merge coverage reports" failures.
Moving the report inline to ci.yml means it runs from the PR branch
and uses the current report template. The report now shows:
- per-platform status (Ubuntu/Windows/macOS columns)
- unified test counts from the single vitest run
- coverage with base branch (main) delta comparison
- commit SHA for traceability
Also removes the save-pr-meta job since the report no longer needs
a separate workflow_run trigger.
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Gergo Magyar <gergomagyar@icloud.com>
|
||
|
Gergő Magyar
|
973c7bfbf0
|
feat: ACCESSES edge type with read/write field access tracking (#372)
* feat: Phase 1 ACCESSES edge type — read tracking from chain resolution Add ACCESSES relationship type to track field read access during call chain resolution. When walkMixedChain resolves a field access (e.g., user.address.save()), an ACCESSES edge with reason 'read' is emitted from the calling function to the Property node. Schema: ACCESSES added to RelationshipType, REL_TYPES, VALID_RELATION_TYPES, context queries, tools/resources descriptions. Excluded from default impact BFS to prevent traversal explosion. Implementation: resolveFieldAccessType now returns FieldResolution with fieldNodeId. walkMixedChain accepts optional onFieldResolved callback. makeAccessEmitter factory provides Set-based dedup per source node. Bug fix: Added Java 'field_access' to FIELD_ACCESS_NODE_TYPES — was missing, causing extractMixedChain to fail for Java member access. * feat: Phase 2 ACCESSES write edges — assignment detection across 12 languages Add tree-sitter query patterns for field write detection (obj.field = value) across all supported languages: TS/JS, Python, Java, Go, C++, C#, Rust, PHP, Ruby (setter syntax), Kotlin, Swift. Processing: Sequential path handles assignment captures inline. Worker path extracts ExtractedAssignment data for deferred resolution via new processAssignmentsFromExtracted function. Bug fix: Kotlin/Swift assignment queries used invalid navigation_expression wrapper — fixed to match actual directly_assignable_expression AST structure. Tests: Write access integration tests for TS, Java, Python, Go with dedicated fixtures. All use strict toBe() assertions. * test: add unit tests for call-routing, shared type extractors, and symbol-table branches Add 215 new unit tests across 3 files to increase branch coverage toward the 23% global threshold (was 21.49%): - call-routing.test.ts (49 tests): Ruby call routing — require/require_relative, include/extend/prepend heritage, attr_accessor properties with YARD types - shared-type-extractors.test.ts (108 tests): pure string functions — extractElementTypeFromString, stripNullable, extractReturnTypeName, methodToTypeArgPosition, getContainerDescriptor - symbol-table.test.ts (+29 tests): Property/fieldByOwner index, metadata spread branches, lazy callable index, lookupExactFull shape * fix: defer write-access resolution to fix Ruby cross-file property timing Ruby attr_accessor properties are registered during processCalls (not the parsing phase), so lookupFieldByOwner fails when service.rb is processed before models.rb. Fix by collecting pending write-access edges during the file loop and resolving them after all files are done. Also adds write-access integration tests and fixtures for 7 languages (C++, C#, JS, Kotlin, PHP, Ruby, Rust), Ruby compound assignment query, PHP static property write query, and Kotlin property type extraction. * fix: address PR #372 review — write-access constructor bindings parity and docs - Add verified constructor bindings fallback to write-access resolution in both sequential path (receiverIndex lookup) and worker path (constructorBindings param for processAssignmentsFromExtracted), closing the read/write ACCESSES edge asymmetry for factory-returned receivers - Clarify inner guard control flow comment in processCalls match loop - Document Go inc_statement/dec_statement gap in roadmap - Clarify PHP nullsafe write footnote (invalid syntax, not just untracked) - Update symbol-table tests for intentional fieldByOwner behavior change (Properties without declaredType now indexed for dynamic language write-access tracking) |