check-style
Skill2 starsChanged 4 days agoApache-2.0View on GitHub
Use to judge a Keel entry against the style guide - a threat or a mitigation card - after writing or changing one, or when auditing the catalog's content. Not for structural checks (schema, links, vocabularies), which are `keel validate`.
In this repository13 other files it gives its agents.
- assess-genai-with-librarySkill
- assessing-genai-securitySkill
- check-styleSkill
- folding-into-the-modelSkill
- humanizerSkill
- keel-designSkill
- tighten-textSkill
- assess-genai-with-librarySkill
- assessing-genai-securitySkill
- folding-into-the-modelSkill
- humanizerSkill
- keel-designSkill
- tighten-textSkill
The file
Files
check-style
Judges the content of an entry against the style guide: the per-field bar, and the few things about a record that no single field's bar can hold.
This skill applies the bar, it does not own it. If a field is wrong and the bar is silent, the finding is against the bar, and the fix is update_style_guide, not a patch to the entry. Re-encoding rules here would create a second source of truth that drifts.
Judgment, not structure. The deterministic gate - schema, links, vocabularies, rules - is uv run keel validate and check_library_health. Assume it passed. It cannot read a sentence, which is the whole reason this exists.
Two modes
One entry, after a write. The default, and the one that matters. Run it on whatever you just created or changed, in the same turn, before reporting it as done. No agents, no report file: read the entry, apply rubric.md, give the verdict in the reply.
The catalog, as an audit. Dispatch review agents in parallel, four or five entries each, same rubric. Write the collected result using report-template.md, to a path the caller names.
Procedure
Both modes run rubric.md. In short:
- Fetch the bar.
get_style_guide(entity_type="threat")plusweaknessandmitigation_link, orget_style_guide(entity_type="mitigation")plusimplementation. - Rate every populated field PASS / MINOR / FAIL against its own bar.
- Apply the record-level checks in rubric section B - the ones that need reading the whole entry.
ENTRY_ID | field | PASS/MINOR/FAIL | current | expected (per bar) | reasoncurrent and expected are not optional. A finding without both cannot be acted on by anyone but its author.
Severity
| Criteria | |
| critical | The record is the wrong shape: an umbrella covering chains closed by different controls, a card named after the threat that prompted it, a technique or a consequence sitting in weakness, reachability that re-negates the weakness or names a control, a soft link claiming it closes the threat, a card with no evidence behind a claim that needs it |
| major | A vague or non-architectural weakness, a wrong nature, an attacker-only source on a threat that is real without one, thin reachability, scope written against one threat, org-specific prose in a card field, acceptance criteria that no reviewer could check |
| minor | Wording, title phrasing, tags, reference relevance |
Common mistakes
| Mistake | Fix |
| Re-encoding rules in the review | Fetch them from the style guide; if a rule is missing, the finding is against the bar |
| A "good enough" verdict | PASS or FAIL. MINOR is a defect, not a hedge |
No current or expected | Always both, or the finding is unactionable |
| Judging only the fields | The record-level checks in rubric B are where the expensive mistakes are |
| Passing a fluent card | Fluency is what a card written from memory has. Check that the evidence exists and supports what the card claims |
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.

