aiwf-sync
Skill1 starsChanged 3 months agoMITView on GitHub
Check whether docs/AI_WORKFLOW.md is in sync with the AI Workflow implementation — routes, phase/skill table, config shape, card format, and 'Where it lives' file inventory. Use after any change to the AIWF connection, or to audit before a release. Reports concrete drift findings; optionally fixes them.
In this repository6 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- config-fieldSkill
- docs-updateSkill
- hangar-validateSkill
- releaseSkill
- smokeSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. About 0 are common, 5 are distinctive, and 10 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
10
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 16
Read docs/AI_WORKFLOW.md in full.
- Very specific: line 20
Read server/src/aiwf.ts. Find SKILL_GROUPS (the phase → skills mapping) and DEFAULT_COLUMNS.
- Very specific: line 24
Every phase in SKILL_GROUPS appears in the table
- Distinctive: line 25
Every skill listed under each phase matches
- Very specific: line 30
Read server/src/index.ts. Find every route registered under /api/aiwf/ (lines with app.get, app.post, app.patch, app.delete where the path …
- Distinctive: line 35
Every route in the code appears in the table
- Distinctive: line 36
No routes in the table are absent from the code
- Very specific: line 41
Read server/src/config.ts.
- Very specific: line 44
Read server/src/types.ts. Find AiwfProject.
- Very specific: line 48
Every field of AiwfProject that is required/meaningful is shown in the example
- Very specific: line 53
Read server/src/aiwf.ts. Find the card create/read logic (look for frontmatter field names like key, title, status, kind, skill, pr, …
- Distinctive: line 59
Every frontmatter field mentioned in the doc exists in the code
- Very specific: line 82
Read the install/uninstall section in server/src/aiwf.ts (detectAiwf, installAiwf, uninstallAiwf).
- Very specific: line 93
If no drift is found anywhere, report "docs/AI_WORKFLOW.md is in sync with the implementation."
- Distinctive: line 95
If drift is found, ask: "Fix these now?"
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Audit docs/AI_WORKFLOW.md against the live AIWF implementation and report (or fix) drift.
The rule in CLAUDE.md: whenever you change the AI Workflow connection — its routes, board model,
phases/skills, config shape, install flow, or card format — update docs/AI_WORKFLOW.md in the same
change. This skill makes that rule checkable.
Steps
1. Read the doc
Read docs/AI_WORKFLOW.md in full.
2. Check: phase/skill table
Read server/src/aiwf.ts. Find SKILL_GROUPS (the phase → skills mapping) and DEFAULT_COLUMNS.
Compare against the phase/skill table in the doc (## What aiwf is section). Check:
- Every phase in
SKILL_GROUPSappears in the table - Every skill listed under each phase matches
DEFAULT_COLUMNSmatches the board column sequence shown in the doc
3. Check: API route table
Read server/src/index.ts. Find every route registered under /api/aiwf/ (lines with
app.get, app.post, app.patch, app.delete where the path starts with /api/aiwf).
Compare against the API table in the doc (## API section). Check:
- Every route in the code appears in the table
- No routes in the table are absent from the code
- HTTP methods match
4. Check: config shape example
Read server/src/config.ts. Find the validateConfig block that handles aiWorkflow and
the cleanAiwfProjects function.
Read server/src/types.ts. Find AiwfProject.
Compare against the JSON example under ## Configuration in the doc:
- Every field of
AiwfProjectthat is required/meaningful is shown in the example - The validation rules described in the doc (id/name/repoPath required, columns optional) match the code
5. Check: card file format
Read server/src/aiwf.ts. Find the card create/read logic (look for frontmatter field names like
key, title, status, kind, skill, pr, archived). Also find the HANGAR_HISTORY
comment block pattern.
Compare against ### Card file format in the doc:
- Every frontmatter field mentioned in the doc exists in the code
- The HANGAR_HISTORY block syntax matches
6. Check: "Where it lives" inventory
The ## Where it lives section lists source files for Server and Web. Check that each listed
file still exists at the stated path:
server/src/aiwf.tsserver/src/index.tsserver/src/sessions.tsserver/src/config.tsserver/src/types.tsweb/src/components/AiWorkflow.tsxweb/src/App.tsxweb/src/api.ts,web/src/types.tsserver/src/__tests__/aiwf.test.tsserver/src/__tests__/index.aiwf.test.ts
Flag any that are missing or have moved.
7. Check: install/uninstall flow
Read the install/uninstall section in server/src/aiwf.ts (detectAiwf, installAiwf,
uninstallAiwf). Confirm the detection logic (bin path and/or skills presence) and the
install/uninstall commands match what the doc says under ## Install / detect / uninstall.
8. Report findings
For each section, report:
- In sync — no drift found
- Drift — describe the specific mismatch (e.g. "Route
PATCH /api/aiwf/projects/:idis in the code but missing from the API table")
If no drift is found anywhere, report "docs/AI_WORKFLOW.md is in sync with the implementation."
If drift is found, ask: "Fix these now?"
- Yes → update
docs/AI_WORKFLOW.mdto match the code (code is the source of truth), then runnpm run typecheckto confirm no accidental edits broke anything. - No → leave the doc as-is and just report the findings.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

