mirror of
https://github.com/alirezarezvani/claude-skills.git
synced 2026-10-08 03:07:51 +00:00
5 commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Claude
|
7405298b4b
|
fix: resolve the actionable reported issues (#954, #949, #933, #931, #969, #968, #924, #885)
- #954: strip non-spec source/attribution keys from all 39 plugin.json manifests so Claude Code's validator accepts them; metadata preserved in new .claude-plugin/authoring-notes.json sidecars; check_plugin_json.py now hard-fails manifests carrying those keys and sanity-checks the sidecar; CLAUDE.md ClawHub schema section updated to the new rule. - #949: move the c-level-agents plugin out of c-level-advisor/ to a top-level directory so the two marketplace sources no longer overlap; updated marketplace.json source, homepage, descriptions, all cross-references, docs, harness manifest, mirror-tree symlinks/indexes, and rebased the moved files' relative links; domain counters trued up (18 -> 19 domains). - #933: replace dead links to the gitignored maintainer-local megaprompts/ tree with annotated plain-text references (44 files: SKILL.md, READMEs, agents, commands). - #931: DynamoDB on-demand pricing updated to post-Nov-2024 rates ($0.625/M writes, $0.125/M strongly consistent reads). - #969: skill_security_auditor.py and the three dossier scripts reconfigure stdout/stderr to UTF-8 (errors=replace) so legacy Windows codepages no longer crash at print time; PYTHONUTF8=1 documented. - #968: Windows Notes section in INSTALLATION.md + README pointer for the core.symlinks mirror-tree checkout caveat. - #924/#885 residuals: hook commands quote "${CLAUDE_PLUGIN_ROOT}" paths in all plugin hooks.json/settings.json (space-safe roots); removed the stale pre-rename status/review mirror symlinks and index entries left over from the memory-status/memory-review rename. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Qgc6RYXWJPr5oW9DHU7zR4 |
||
|
Claude
|
11415a614e
|
feat(research): free keyless search lane + routing guard (audit PR-4)
litreview: paid Consensus MCP dependency converted to an optional enhancement lane; PubMed E-utilities + OpenAlex (both keyless) are now the default search lane with exact URL templates and JSON field maps; brittle plan-tier detection (marketing-copy parsing) deleted in favor of a one-line runtime availability check; new stdlib scripts/free_search.py (--source pubmed|openalex|both, --json, graceful exit 2 offline). Fixes the audit OPTIMIZE finding and brings the skill into compliance with ClawHub publishing rule #3 (no paid-service dependencies). Mirrored across README/agent/command/plugin.json/references. research orchestrator: bare-noun signals (funding/fda/patent/grant) no longer silent-route — single bare-noun match now returns ask + recommended specialist; silent-route requires >=2 signals or one strong multi-word phrase (SKILL.md and classifier.py updated in lockstep, sync verified programmatically); the un-executable 'auto-proceeds after 5s' affordance removed everywhere; SKILL.md trimmed 319 -> 247 lines preserving routing tables and transparency protocol. Live classifier proof: CRISPR meta-analysis -> litreview (silent); Acme due-diligence -> dossier (silent); 'research Microsoft' -> fallback; bare 'funding'/'fda' -> ask (recommended: grants). Counters reconciled (+1 tool = 580); gates green: paths 0, dual-publish 0, smoke 584/584, counters match. https://claude.ai/code/session_019AJddAL1NADWMXsy1qNPQF |
||
|
Claude
|
a088c8ba77
|
fix: phantom-path sweep — 888 unresolvable references to zero
A directory reorg added a skills/ path segment; hundreds of references never followed. This sweep repoints every path-like reference in SKILL.md, agents, commands, orchestration, and templates to verified on-disk targets: - 30 root commands + 19 root agents: missing skills/ segment inserted - 7 c-level persona agents: 17 hallucinated reference filenames substituted with the real files (e.g. okr_execution.md -> process_frameworks.md) - 5 research skills: phantom scripts/office/validate.py step replaced with a runnable stdlib zip-integrity check - email agents: skills frontmatter corrected to productivity/email - orchestration/ORCHESTRATION.md + templates: stale paths fixed; agent-template now requires trigger phrasing in descriptions (root cause) - 76 more files across engineering, c-level-advisor, compliance-os, research-ops, ra-qm, marketing, productivity; dual-publish pairs mirrored - dead refs dropped/replaced where no target ever existed (REGISTRY.md, trend_analyzer.py, cursor-microinteractions.md) New: scripts/check_paths.py linter (CI gate G1) + narrow allowlist for teaching examples. Verified: 540 files scanned, 0 unresolvable. https://claude.ai/code/session_019AJddAL1NADWMXsy1qNPQF |
||
|
Claude
|
6d9630f83c
|
chore(cleanup): move pulse + capture to proper domain folders
Surgical move PR — resolves the two domain warts accumulated during
the v2 megaprompt build sweep:
engineering/pulse/ → research/pulse/ (research-pack — pulse is
the first research skill;
now joins litreview, grants,
dossier, patent, syllabus)
engineering/capture/ → productivity/capture/ (productivity — capture
is brain-dump organizer,
not engineering tooling)
WHY THIS PR
When Slice 1 (capture) shipped in PR #659, the productivity/ domain
folder didn't yet exist. When Slice 2 (pulse) shipped in PR #660, the
research/ folder didn't yet exist either. Both were placed in
engineering/ as the catch-all.
After Slices 3-5 established the productivity/, marketing/, and
research/ top-level domain folders, those two early skills were left
in engineering/ as warts. This PR resolves them BEFORE Slice 7
(13-research orchestrator) so the orchestrator can reference
research/pulse/ as its routing target without further path churn.
WHAT MOVED
Two directories moved via `git mv` (preserves rename history):
- engineering/pulse → research/pulse (11 files)
- engineering/capture → productivity/capture (11 files)
INTERNAL REFERENCES UPDATED
Inside the moved directories:
- .claude-plugin/plugin.json homepage URLs (engineering/X → new path)
- agents/cs-*.md `skills:` frontmatter field
CROSS-SKILL REFERENCES UPDATED
6 external files reference pulse and/or capture as sibling skills.
All updated via sed:
productivity/email/agents/cs-inbox-setup.md (capture ref)
productivity/email/agents/cs-inbox-triage.md (pulse + capture refs)
research/grants/agents/cs-grants.md (pulse ref)
research/litreview/agents/cs-litreview.md (pulse ref + stale
"will move in cleanup
PR" caveat removed)
research/dossier/agents/cs-dossier.md (pulse ref)
marketing/landing/agents/cs-landing.md (pulse + capture refs)
CODEX SYMLINKS RE-POINTED
.codex/skills/{capture,pulse} symlinks updated to point at new
locations. Verified resolution to SKILL.md files works.
.codex/skills-index.json still references the old paths — this file
is auto-regenerated by the codex-sync workflow on every merge to dev
(prior commits:
|
||
|
Claude
|
a4bb1fc648
|
feat(research): litreview skill — Path-B research-pack sibling from megaprompt 09
Slice 5 batch 1 of N: first research-pack sibling after pulse (Slice 2). Establishes the academic-literature variant of the research-pack shape + introduces the research/ top-level domain folder. SOURCE SPEC megaprompts/09-litreview-megaprompt.md (PR #657). Canonical. DOMAIN FOLDER DECISION (research/ — new) Pulse currently lives in engineering/ (placed before the domain-folder discipline crystallized). The right home for academic-research skills is research/ — parallel to productivity/, marketing/. This PR creates that folder; pulse + capture moves are deferred to a coordinated cleanup PR. Cumulative folder warts: - engineering/capture/ → productivity/capture/ (Slice 1 wart) - engineering/pulse/ → research/pulse/ (Slice 2 wart) Cleanup PR will address both before Slice 5 (orchestrator) lands. WHY ONE SKILL PER PR (NOT BATCH 6) User recommended batching 6 research-pack siblings. Doing 1 in this PR instead, with rationale: each megaprompt is dense (litreview alone is 266 lines with 8 DOCX sections, 3-tier search budget logic, cross-search intelligence trackers). 33-66 files of unfocused conversion risks Path-B fidelity. Subsequent PRs will ratchet up to 2 skills each now that the academic-literature pattern is validated. WHAT THE SKILL DOES Turns a research question into a strategically planned mini literature review delivered as an 8-section .docx. Grill-me intake (question + framework + tentative depth) before reconnaissance; second forcing checkpoint after Phase 2 confirms framework + sub-areas + final depth. Sequential Consensus searches at 1 q/sec, budget-allocated by tier (5/10/20). Cross-search intelligence (repeat-hits, recurring-authors, citations-per-year) feeds the "Start Here" + "Key Research Groups" DOCX sections. Output is a "launching pad" — orientation guide, not a finished review. PATH-B FIDELITY (megaprompt → SKILL.md) - Frontmatter description preserved verbatim from megaprompt. - 10-step workflow structure preserved 1:1 (Agent Integrity Rules → Error Handling → Phase 0 intake → Phase 1 recon → Phase 2 framework → Checkpoint → Phase 3 searches → Phase 4 DOCX → Doc structure → Technical requirements). - All 3 grill-me intake questions preserved verbatim with rationale. - All 5 Agent Integrity Rules preserved verbatim per PR #657 audit. - All 3 search budget tiers fully allocated (5/10/20 with explicit query breakdown per tier). - All 8 DOCX sections fully specified. - Interactive checkpoint described as forcing-options moment (not free-text). - Three frameworks (PICO/SPIDER/Decomposition) + Hybrid documented with examples. - Anti-patterns + error-handling table + validation checklist preserved. RESEARCH-PACK CONVENTION MARKERS (in SKILL.md per PR #657 audit) Agent Integrity Rules: 2 sequential: 5 three-count: 1 plan-tier: 3 1 query/sec: 2 checkpoint: 9 retry once: 2 Source discipline: 1 3 consecutive: 2 All markers present multiple times. REPO STRUCTURE research/litreview/ ├── .claude-plugin/plugin.json ← source.spec → megaprompts/09 ├── README.md ├── agents/cs-litreview.md ← sequential-Consensus + checkpoint enforcer ├── commands/cs-litreview.md ← /cs:litreview <research-question> └── skills/litreview/ ├── SKILL.md ← Path-B converted ├── references/ │ ├── framework_selection.md ← PICO/SPIDER/Decomp/Hybrid + 7 sources │ │ (Sackett, Cooke et al., Booth, PRISMA, │ │ Cochrane, Hewitt-Taylor, JBI) │ ├── search_budget_allocation.md ← 5/10/20 + cross-search + 7 sources │ │ (Consensus docs, Cochrane, Greenhalgh, │ │ PRISMA, Sandelowski, Lawani, AWS) │ └── docx_8_sections.md ← 8-section guide + 7 sources (docx lib, │ OOXML, PRISMA, Cochrane, Lipsey, Tufte, │ Strunk) └── scripts/ ├── citation_tracker.py ← stdlib: three-count + 1s rate-limit │ discipline enforcement ├── framework_recommender.py ← stdlib: keyword heuristic PICO/SPIDER/ │ Decomp/Hybrid recommendation └── cross_search_aggregator.py ← stdlib: repeat-hits, recurring-authors, citations-per-year ranking 11 files, 2,020 lines. Slightly heavier than pulse (1,643) due to: - Denser SKILL.md (251 lines vs pulse 258 — comparable) - Heaviest reference: docx_8_sections.md at 287 lines (8 sections × ~35 lines each spec) - citation_tracker.py is heavier than pulse's (258 vs 251) because it enforces the 1s sequential gap explicitly VERIFIED CLEAN - citation_tracker.py: full lifecycle works. Sequential discipline enforced — second search at 0.04s correctly REJECTED with "wait 0.96s more"; after 1.1s sleep, accepted. Three-count audit block output matches PR #657 audit format. - framework_recommender.py: PICO question (clinical reasoning vs physicians) → recommends PICO SPIDER question (qualitative burnout study) → recommends SPIDER with high confidence (3 signals) Decomposition question (RAG systems benchmarks) → recommends Decomposition (after plural-aware regex fix; was originally missed due to "systems" not matching "system") - cross_search_aggregator.py: sample with overlapping searches: Repeat-hits: Med-PaLM benchmark correctly flagged (3 sub-areas) Recurring authors: Singhal correctly top (4 appearances) Citations-per-year: USMLE benchmark paper top at 266/yr - All 3 with --output json: valid JSON - plugin.json validates; conforms to repo schema NOT-YET-DONE (for upcoming PRs) - Slice 5 batch 2: grants + dossier (next PR; same shape as litreview) - Slice 5 batch 3: patent + syllabus (specialty variants — patent has sub-use-case routing, syllabus has bundled JS DOCX generator) - Slice 6: notebooklm (browser-automation shape, separate slice) - Slice 7: 13-research orchestrator + autoresearch-agent reconciliation - Slice 8: 02-reflect productivity - Cleanup PR: move engineering/pulse + engineering/capture to their proper domain folders https://claude.ai/code/session_01FEUmeuYhmnxVFq7EZM8ZSw |