checkpoint

⚒️ Generated with [Fabro](https://fabro.sh)

stages/005-implement@1/diff.patch

@@ -0,0 +1,128 @@
+diff --git a/apps/fabro-web/app/components/automation-form.tsx b/apps/fabro-web/app/components/automation-form.tsx
+index 4788557ec..8213e0c0a 100644
+--- a/apps/fabro-web/app/components/automation-form.tsx
++++ b/apps/fabro-web/app/components/automation-form.tsx
+@@ -86,7 +86,7 @@ export function automationFormValuesFromRun(
+     name,
+     repository,
+     ref:        cloneBranch ?? EMPTY_AUTOMATION_FORM.ref,
+-    workflow:   run.workflow.slug?.trim() || snakeify(workflowName),
++    workflow:   run.workflow.slug?.trim() || kebabify(workflowName),
+   };
+ }
+ 
+@@ -124,14 +124,6 @@ export function kebabify(value: string): string {
+     .replace(/^-|-$/g, "");
+ }
+ 
+-export function snakeify(value: string): string {
+-  return value
+-    .toLowerCase()
+-    .replace(/[^a-z0-9_]+/g, "_")
+-    .replace(/_+/g, "_")
+-    .replace(/^_|_$/g, "");
+-}
+-
+ function firstPresentString(...values: Array<string | null | undefined>): string {
+   for (const value of values) {
+     const trimmed = value?.trim();
+@@ -300,15 +292,15 @@ export function AutomationFormFields({
+         </Row>
+         <Row
+           title={<Label required>Workflow slug</Label>}
+-          help="Snake-case identifier used in the workflow file name (e.g. fix_build.fabro)."
++          help="Dash-separated identifier matching the workflow directory name (e.g. patch-cves)."
+         >
+           <input
+             type="text"
+             name="workflow_slug"
+             aria-label="Workflow slug"
+             value={values.workflow}
+-            onChange={(e) => patch({ workflow: snakeify(e.target.value) })}
+-            placeholder="fix_build"
++            onChange={(e) => patch({ workflow: kebabify(e.target.value) })}
++            placeholder="patch-cves"
+             autoComplete="off"
+             spellCheck={false}
+             className={`${INPUT_CLASS} font-mono`}
+diff --git a/apps/fabro-web/app/routes/automations-new.test.tsx b/apps/fabro-web/app/routes/automations-new.test.tsx
+index 32a88dccc..7c6defc47 100644
+--- a/apps/fabro-web/app/routes/automations-new.test.tsx
++++ b/apps/fabro-web/app/routes/automations-new.test.tsx
+@@ -109,7 +109,7 @@ function makeRun(overrides: Record<string, unknown> = {}) {
+     goal:             "Fix CI",
+     title:            "Fix failing tests",
+     workflow:         {
+-      slug:       "fix_ci",
++      slug:       "fix-ci",
+       name:       "Fix CI",
+       graph_name: "ci_graph",
+       node_count: 0,
+@@ -219,6 +219,18 @@ function fieldValue(renderer: TestRenderer.ReactTestRenderer, label: string) {
+   return renderer.root.findByProps({ "aria-label": label }).props.value;
+ }
+ 
++function changeField(
++  renderer: TestRenderer.ReactTestRenderer,
++  label: string,
++  value: string,
++) {
++  act(() => {
++    renderer.root.findByProps({ "aria-label": label }).props.onChange({
++      target: { value },
++    });
++  });
++}
++
+ function switchChecked(renderer: TestRenderer.ReactTestRenderer, label: string) {
+   const props = renderer.root.findByProps({ "aria-label": label }).props;
+   return props["aria-checked"] ?? props.checked;
+@@ -265,6 +277,16 @@ describe("AutomationsNew", () => {
+     expect(switchChecked(renderer, "Enable scheduled triggers")).toBe(false);
+   });
+ 
++  test("workflow slug input normalizes to kebab-case and preserves dashes", async () => {
++    const { renderer } = await renderAutomationsNew("/automations/new");
++
++    changeField(renderer, "Workflow slug", "patch-cves");
++    expect(fieldValue(renderer, "Workflow slug")).toBe("patch-cves");
++
++    changeField(renderer, "Workflow slug", "Patch CVEs");
++    expect(fieldValue(renderer, "Workflow slug")).toBe("patch-cves");
++  });
++
+   test("/automations/new?from_run=run_1 pre-populates from run and settings data", async () => {
+     currentRun = makeRun();
+     currentRunSettings = makeRunSettings();
+@@ -275,7 +297,7 @@ describe("AutomationsNew", () => {
+     expect(fieldValue(renderer, "Automation slug")).toBe("fix-failing-tests");
+     expect(fieldValue(renderer, "Repository")).toBe("qltysh/fabro");
+     expect(fieldValue(renderer, "Default branch")).toBe("feature/from-run");
+-    expect(fieldValue(renderer, "Workflow slug")).toBe("fix_ci");
++    expect(fieldValue(renderer, "Workflow slug")).toBe("fix-ci");
+     expect(switchChecked(renderer, "Enable manual and API triggers")).toBe(true);
+     expect(switchChecked(renderer, "Enable scheduled triggers")).toBe(false);
+     expect(
+@@ -285,6 +307,22 @@ describe("AutomationsNew", () => {
+     expect(queryCalls).toContainEqual({ hook: "useRunSettings", id: "run_1" });
+   });
+ 
++  test("/automations/new?from_run=run_1 kebab-cases workflow name fallback", async () => {
++    currentRun = makeRun({
++      workflow: {
++        slug:       "",
++        name:       "Patch CVEs",
++        graph_name: "PatchCves",
++        node_count: 0,
++        edge_count: 0,
++      },
++    });
++
++    const { renderer } = await renderAutomationsNew("/automations/new?from_run=run_1");
++
++    expect(fieldValue(renderer, "Workflow slug")).toBe("patch-cves");
++  });
++
+   test("missing source run data renders an editable empty form with a non-blocking error", async () => {
+     currentRun = null;
+     currentRunError = new Error("not found");

stages/005-implement@1/status.json

@@ -0,0 +1,6 @@
+{
+  "outcome": "succeeded",
+  "notes": "Stage completed: implement",
+  "failure_reason": null,
+  "timestamp": "2026-07-07T13:26:27.809284893Z"
+}

stages/006-simplify_fable@1/prompt.md

@@ -0,0 +1,241 @@
+Goal: # Plan: automation "Workflow slug" field must use kebab-case (dashes), not snake_case
+
+## Summary
+
+The web UI's New/Edit Automation form silently rewrites the **Workflow slug**
+field to snake_case on every keystroke (`patch-cves` → `patch_cves`). The rest
+of the system uses kebab-case (dashes) for workflow slugs, and workflow
+directories on disk are dash-separated (`.fabro/workflows/patch-cves/`). So a
+snake_cased selector points at a non-existent directory, and scheduled
+automations targeting any multi-word workflow **silently never fire**.
+
+Fix: change the Workflow slug field (and the "create from run" fallback) to use
+the form's existing `kebabify()` helper instead of `snakeify()`, update the
+field copy, remove the now-unused `snakeify()`, and update the test. This is a
+**frontend-only** change; the backend already accepts dashes and needs no edit.
+
+## Root cause (verified, with file:line)
+
+All in `apps/fabro-web/app/components/automation-form.tsx`:
+
+- `snakeify()` (lines 127-133) — `.toLowerCase().replace(/[^a-z0-9_]+/g, "_")…`
+  converts any dash to an underscore.
+- The Workflow slug input (lines 302-311) applies it on every keystroke:
+  ```tsx
+  // line 303 help text:
+  help="Snake-case identifier used in the workflow file name (e.g. fix_build.fabro)."
+  // line 310:
+  onChange={(e) => patch({ workflow: snakeify(e.target.value) })}
+  // line 311:
+  placeholder="fix_build"
+  ```
+- The "create automation from run" prefill (line 89) also snake-cases the
+  fallback name:
+  ```tsx
+  workflow: run.workflow.slug?.trim() || snakeify(workflowName),
+  ```
+- `kebabify()` (lines 119-126) already exists and is used for the automation's
+  own id field (lines 85, 214, 220). It is the correct normalizer to reuse.
+
+`snakeify` is referenced in exactly two places (lines 89 and 310); once both
+move to `kebabify`, the `snakeify` definition (lines 127-133) is dead code.
+
+## Why kebab-case is correct (justification)
+
+The snake_case rule is a UI-only outlier. Every other layer treats workflow
+slugs as kebab-case:
+
+- Backend `validate_workflow_selector` (`lib/crates/fabro-automation/src/model.rs:366`)
+  allows both `-` and `_`, so it will accept a dashed value with no change.
+- Real workflow directories are dash-separated: `patch-cves`, `gh-triage`,
+  `implement-plan`, `card-game`, `implement-issue`, etc. Only single-word ones
+  (`smoke`, `hello`) survive `snakeify` unchanged.
+- `workflow_slug_from_path` (`lib/crates/fabro-config/src/project.rs:111`)
+  preserves dashes, and `lib/crates/fabro-workflow/src/run_options.rs:31`
+  documents the slug as the **literal workflow directory name**.
+- `AutomationId` (`lib/crates/fabro-automation/src/id.rs`) actually **forbids
+  underscores** — dashes only.
+
+So kebab is the established convention; the form is simply wrong. No backend
+change is needed or wanted.
+
+### Two distinct names — do not conflate (important)
+
+A workflow has two separate identifiers, and only one is relevant here:
+
+- **Workflow slug** — derived from the *directory* name, kebab-case
+  (e.g. `patch-cves` from `.fabro/workflows/patch-cves/`). This is what the
+  automation `workflow` selector resolves against, and what this fix must
+  produce. (Seen as `workflow_slug: "patch-cves"` on runs.)
+- **Graph name** — the identifier in the DOT source, `digraph PatchCves { … }`,
+  typically CamelCase. This is internal graph identity (`name: "PatchCves"` on
+  runs); it is **not** used to resolve the automation target.
+
+The Workflow slug field must be the kebab directory slug. Do **not** change it
+toward the CamelCase graph name, and do not touch graph names anywhere. This
+also explains the original mistake: the author treated the file/dir name as
+snake_case when it is actually kebab-case.
+
+## Failure mode this fixes (for context)
+
+1. User types `patch-cves` → form stores `patch_cves`.
+2. Server stores it verbatim (no normalization; validator accepts underscores).
+3. On each cron tick, `automation_materializer` resolves the selector literally
+   → looks for `.fabro/workflows/patch_cves/` → `WorkflowNotFound`
+   (`lib/crates/fabro-server/src/automation_materializer.rs` ~219-229).
+4. `lib/crates/fabro-server/src/server/automation_scheduler.rs:221-239` logs
+   `"Failed to materialize scheduled automation run"` and returns — no run
+   created, waits for the next tick. The automation appears to do nothing.
+
+## Where it was introduced (provenance, informational)
+
+- `snakeify` + the Workflow slug field + "Snake-case identifier…" help text:
+  commit `2e85a1ec4` "Add New Automation form and refresh Secrets form layout"
+  (2026-05-24). Present from the form's first draft; no commit message explains
+  why the workflow field is snake-cased.
+- Carried into the shared component: `87516c25c` (2026-05-28).
+- Prefill fallback snake-cases the name: `a65473f21` / PR #454 (2026-05-29).
+- Backend validator that (correctly) allows dashes: `e13de9faa` / PR #428.
+
+## Implementation steps
+
+Frontend only. Use red/green TDD.
+
+1. **Update the test first** — `apps/fabro-web/app/routes/automations-new.test.tsx`:
+   - The assertion at ~line 278 currently expects the snake output
+     `expect(fieldValue(renderer, "Workflow slug")).toBe("fix_ci")`. Change the
+     expected value to the kebab form (`"fix-ci"`) to match the new behavior.
+   - Add a regression assertion: typing `patch-cves` into the "Workflow slug"
+     field leaves it as `patch-cves` (dashes preserved, NOT converted to
+     underscores). Also confirm `Patch CVEs` → `patch-cves`.
+   - Run the test and confirm it fails against current code (red).
+
+2. **Fix the field** — `apps/fabro-web/app/components/automation-form.tsx`:
+   - Line 310: `onChange={(e) => patch({ workflow: snakeify(e.target.value) })}`
+     → use `kebabify(e.target.value)`.
+   - Line 89: `run.workflow.slug?.trim() || snakeify(workflowName)`
+     → `run.workflow.slug?.trim() || kebabify(workflowName)`.
+   - Line 303 help text → describe dash-separated, e.g.
+     `"Dash-separated identifier matching the workflow directory name (e.g. patch-cves)."`
+   - Line 311 placeholder `"fix_build"` → `"patch-cves"` (or `"fix-build"`).
+   - Remove the now-unused `snakeify()` function (lines 127-133). Keep
+     `kebabify()` (still used for the id field and now the workflow field).
+   - Verify no references remain: `grep -rn "snakeify" apps/fabro-web/app`
+     should return nothing.
+
+3. **Verify** in `apps/fabro-web`:
+   - `bun run typecheck` passes.
+   - `bun test` passes (the updated + new assertions go green).
+
+## Files to touch
+
+- `apps/fabro-web/app/components/automation-form.tsx` — the fix (lines 89, 303,
+  310, 311; remove 127-133).
+- `apps/fabro-web/app/routes/automations-new.test.tsx` — update the snake
+  assertion (~278) and add the dash-preservation regression test.
+- No other files. `automations-new.tsx` / `automations-edit.tsx` submit
+  `values.workflow.trim()` unchanged and need no edit.
+
+## Acceptance criteria
+
+- Typing `patch-cves` in "Workflow slug" yields `patch-cves` (not `patch_cves`).
+- Typing `Patch CVEs` yields `patch-cves`.
+- The "create from run" prefill uses the run's dashed slug when present and
+  kebab-cases the fallback name.
+- Help text and placeholder reflect dash-separated slugs.
+- No `snakeify` references remain (dead code removed).
+- `bun run typecheck` and `bun test` pass in `apps/fabro-web`.
+- No backend/Rust changes.
+
+## Out of scope
+
+- **Backend changes.** `validate_workflow_selector` already accepts dashes;
+  leave it. Do not add snake↔dash conversion on the server.
+- **`kebabify()` itself** — it is correct for the id field; do not modify it.
+- **Migrating already-created automations** whose stored `workflow` is already
+  snake_cased. That is an operational fix (edit the automation via
+  `PUT /api/v1/automations/{id}` with the corrected dashed `workflow`), not part
+  of this code change.
+- The unrelated sandbox `glob` consistency fix (tracked in a separate plan).
+
+## Design decision (stated so it isn't re-litigated)
+
+Keep normalizing the field — do not make it freeform. Normalize to **kebab-case**
+(reusing the existing `kebabify`) so it matches the id field's behavior, the
+backend, and the on-disk workflow directory convention.
+
+## Conventions to follow
+
+- Match the surrounding TypeScript/React style in `apps/fabro-web`.
+- This is an `onChange`/string-helper change only — no new React effects (the
+  repo avoids direct `useEffect`; none is needed here).
+- Keep the change minimal and focused; do not refactor unrelated form code.
+
+
+## Completed stages
+- **toolchain**: succeeded
+  - Script: `command -v cargo >/dev/null || { curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && sudo ln -sf $HOME/.cargo/bin/* /usr/local/bin/; }; cargo --version 2>&1`
+  - Output:
+    ```
+    cargo 1.95.0 (f2d3ce0bd 2026-03-21)
+    ```
+- **preflight_compile**: succeeded
+  - Script: `cargo check -q --workspace 2>&1`
+  - Output: (empty)
+- **preflight_lint**: succeeded
+  - Script: `cargo +nightly-2026-04-14 clippy -q --workspace --all-targets -- -D warnings 2>&1`
+  - Output: (empty)
+- **implement**: succeeded
+  - Model: gpt-5.5, 399.4k tokens in / 7.2k out
+
+
+# Simplify: Code Review and Cleanup
+
+Review all changes for reuse, quality, and efficiency. Fix any issues found. Feel free to use any sub agents you need.
+
+## Phase 1: Identify Changes
+
+Run git diff (or git diff HEAD if there are staged changes) to see what changed. If there are no git changes, review the most recently modified files that the user mentioned or that you edited earlier in this conversation. (You may already have the changes in context, if so, feel free to skip this part)
+
+## Phase 2: Launch Three Review Agents in Parallel
+
+Use the Agent tool to launch all three agents concurrently in a single message. Pass each agent the full diff so it has the complete context.
+
+### Agent 1: Code Reuse Review
+
+For each change:
+
+1. Search for existing utilities and helpers that could replace newly written code. Use Grep to find similar patterns elsewhere in the codebase — common locations are utility directories, shared modules, and files adjacent to the changed ones.
+2. Flag any new function that duplicates existing functionality. Suggest the existing function to use instead.
+3. Flag any inline logic that could use an existing utility — hand-rolled string manipulation, manual path handling, custom environment checks, ad-hoc type guards, and similar patterns are common candidates.
+
+Note: This is a greenfield app, so focus on maximizing simplicity and don't worry about changing things to achieve it.
+
+### Agent 2: Code Quality Review
+
+Review the same changes for hacky patterns:
+
+1. Redundant state: state that duplicates existing state, cached values that could be derived, observers/effects that could be direct calls
+2. Parameter sprawl: adding new parameters to a function instead of generalizing or restructuring existing ones
+3. Copy-paste with slight variation: near-duplicate code blocks that should be unified with a shared abstraction
+4. Leaky abstractions: exposing internal details that should be encapsulated, or breaking existing abstraction boundaries
+5. Stringly-typed code: using raw strings where constants, enums (string unions), or branded types already exist in the codebase
+
+Note: This is a greenfield app, so be aggressive in optimizing quality.
+
+### Agent 3: Efficiency Review
+
+Review the same changes for efficiency:
+
+1. Unnecessary work: redundant computations, repeated file reads, duplicate network/API calls, N+1 patterns
+2. Missed concurrency: independent operations run sequentially when they could run in parallel
+3. Hot-path bloat: new blocking work added to startup or per-request/per-render hot paths
+4. Unnecessary existence checks: pre-checking file/resource existence before operating (TOCTOU anti-pattern) — operate directly and handle the error
+5. Memory: unbounded data structures, missing cleanup, event listener leaks
+6. Overly broad operations: reading entire files when only a portion is needed, loading all items when filtering for one
+
+## Phase 3: Fix Issues
+
+Wait for all three agents to complete. Aggregate their findings and fix each issue directly. If a finding is a false positive or not worth addressing, note it and move on — do not argue with the finding, just skip it.
+
+When done, briefly summarize what was fixed (or confirm the code was already clean).

stages/006-simplify_fable@1/provider_used.json

@@ -0,0 +1,6 @@
+{
+  "mode": "agent",
+  "provider": "anthropic",
+  "model": "claude-fable-5",
+  "reasoning_effort": "xhigh"
+}

stages/006-simplify_fable@1/response.md

@@ -0,0 +1,18 @@
+Review complete and fixes applied. Summary:
+
+## Review verdict
+The implement-stage change was fundamentally sound: all three agents confirmed `snakeify` was removed with zero remaining references repo-wide, `kebabify` reuse is correct, the new help text matches the on-disk kebab-case workflow directory convention, the per-keystroke `kebabify` cost is identical to what it replaced, and the `changeField` test helper follows established suite conventions with proper renderer cleanup.
+
+## Fixed (3 issues)
+1. **Dead export** — `kebabify` in `automation-form.tsx:119` was exported with no importers anywhere (same smell as the just-deleted `snakeify`). Made it module-private.
+2. **Route mount for pure-function coverage** (flagged by two agents) — the "kebab-cases workflow name fallback" test built the full memory-router + mocked-hooks render just to assert the fallback branch of `automationFormValuesFromRun`. The prefill *wiring* is already covered by the adjacent "pre-populates from run" test, so it's now a direct call to the exported pure function (test runtime dropped from ~9ms to 0.07ms).
+3. **Locator duplication** — `renderer.root.findByProps({ "aria-label": label })` appeared 3× across `fieldValue`/`changeField`/`switchChecked`; extracted a shared one-line `byLabel` helper.
+
+## Noted, intentionally skipped
+- **Per-keystroke normalization strips trailing dashes/spaces while typing** (`kebabify("patch-")` → `"patch"`), so multi-word slugs must be pasted or edited mid-string. This is pre-existing, identical to the Automation slug field, and the plan explicitly locks the live-normalization design — worth a separate follow-up (normalize on blur) covering both fields.
+- **`automations.tsx:58-74`** — `slugIconMap`/`slugColorMap` are keyed by snake_case workflow slugs (`fix_build`, …) that kebab-created automations will never match, falling back to the default icon/color. Cosmetic, pre-existing, and outside this plan's "no other files" scope.
+- Hoisting `changeField` into `test-utils.tsx` — only one consumer today; four other test files inline the pattern, so a shared helper is a follow-up, not a requirement.
+
+## Verification
+- `bun run typecheck` — clean
+- `bun test` — 652 pass / 0 fail across 82 files