agentleFS
Iabstergo1Iabstergo1/pdf-to-study-kb/.agents/skills/wiki-lint-semantic/SKILL.md

wiki-lint-semantic

Skill2 starsChanged 10 days agoMITView on GitHub

Run the judgement-requiring health checks on the KB — whether a comparison truly covers the key difference dimensions (L4), whether cross-page conclusions contradict, whether recent kb-save output adds real value (Q2) — emitting Review-Queue proposals only, never editing any wiki page. Use when the user says "do a semantic health check / check for contradictions / see whether the comparison pages are complete". Deterministic lint (L1/L2/L3/L5/L6/broken-links/duplicates) is handled by scripts/pipeline.py lint, not here.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Follow-up handling goes to kb-review: this skill only files proposals.

    line 33
  2. Distinctive:

    No wiki content page edited directly.

    line 68

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

wiki-lint-semantic — semantic lint (the half the finishing CLI cannot do)

Deterministic lint (L1/L2/L3/L5/L6/broken-links/duplicates) is handled by python scripts/pipeline.py lint; this skill does only the judgement-requiring part. It emits Review-Queue proposals only and never directly edits any wiki content page. The execution layer is scripts/pipeline.py; this skill only orchestrates.

1. Triggers / Non-triggers

  • Triggers: "do a semantic health check", "check for contradictions", "see whether the comparison pages are complete", "check whether the kb-save output adds value".
  • Non-triggers: broken links / missing sections / bare evidence IDs / orphan pages and other deterministic lint (use python scripts/pipeline.py lint); read-only Q&A (use kb-query); handling an existing Review-Queue (use kb-review); coverage, Q-chain, formula/evidence spot-checks and other broad QA/audit requests belong to kb-qa — do not hijack its triggers.

2. Inputs

  • wiki/comparisons/**, wiki/topics/**, wiki/synthesis/**, wiki/overview.md, and the relevant concept/lesson/source pages.
  • Recent kb-save query-sessions (especially decision.md, candidate_write_set.json, evidence_refs.json).
  • The existing wiki/Review-Queue/, to avoid duplicate proposals.

3. Outputs

  • wiki/Review-Queue/semantic-lint-<YYYY-MM-DD>.md.
  • Each proposal carries: page path, problem type (L4 / contradiction / Q2), evidence, suggested fix direction, whether a user decision is needed.
  • It does not directly edit comparison/topic/synthesis/concept/lesson/source content pages.

4. Dependencies

  • CLI: may run/reference python scripts/pipeline.py lint --source <source_id> for the deterministic result, but does not re-implement L1/L2/L3/L5/L6/broken-link/duplicate rules.
  • Protocols: docs/skill-runtime/schema.md (page roles / per-type frontmatter contract), save-back-policy.md (Q2 save-value judgement).
  • Follow-up handling goes to kb-review: this skill only files proposals.

5. Persisted artifacts

  • wiki/Review-Queue/semantic-lint-<YYYY-MM-DD>.md: the semantic-lint proposals.
  • A chat summary: scope checked, number of findings, suggestion to handle via kb-review.

6. CLI commands

python scripts/pipeline.py lint --source <source_id>
python scripts/pipeline.py status

These only confirm the deterministic gate / state background; semantic judgements become Review-Queue proposals, never a direct promote or content edit.

7. Workflow

Sub-unitInputOutputAcceptancePersistedFailure stop
L1 scoperequest + vault indexpage set to checkfocus on comparison/contradiction/Q2, not broad QAcheck listvault missing
L2 comparison L4comparison pageskey-dimension gapscovers assumptions/conditions/outcomes/costsproposal draftpage evidence thin
L3 contradictionrelated concept/topic/lessonconflicting claim pairsclaims locatable under one concept/modelproposal draftsource not locatable
L4 Q2kb-save output + evidenceadded-value judgementdistinguishes new synthesis from restatingproposal draftevidence missing
L5 write proposalsproposal draftssemantic-lint fileeach has path/problem/evidence/fix directionReview-Queueduplicate proposal

8. Failure stops / recovery

vault or index missing; deterministic lint already failed and blocks the semantic scope; a page lacks evidence to judge; the finding already has an open proposal; the user asks to edit content pages directly (switch to kb-review and wait for confirmation). Recovery: proposals are the durable output; hand off to kb-review.

9. Acceptance criteria

  • No wiki content page edited directly.
  • semantic-lint-<YYYY-MM-DD>.md written to Review-Queue.
  • Each proposal has a page path, problem description, evidence, and a suggested fix direction.
  • L4/Q2/contradiction judgements do not duplicate deterministic-lint responsibilities.
  • The user was prompted to handle the proposals via kb-review.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.