check-ok
Skill1 starsChanged 2 months agoMITView on GitHub
Check OK — ad-hoc honesty check for any work (roadmap or side research). Scaffolds a side-check freeze if needed, then runs the same freeze-review and optional build-verification path used by every Overseer consumer. Works in Cursor, Claude Code, and via CLI for Copilot / any assistant.
In this repository26 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .cursor/rules/build-verification-required.mdcCursor rule
- .cursor/rules/check-ok-thinking.mdcCursor rule
- .cursor/rules/governance-sync.mdcCursor rule
- .cursor/rules/l1-checkpoints-consumer-owned.mdcCursor rule
- .cursor/rules/no-docs-only-pr-to-main.mdcCursor rule
- .cursor/rules/orchestrator.mdcCursor rule
- .cursor/rules/print-next-closeout.mdcCursor rule
- .cursor/rules/tier-authority.mdcCursor rule
- .cursor/rules/workspace-authority.mdcCursor rule
- build-verification-reviewSkill
- check-okSkill
- deploy-verification-reviewSkill
- freeze-review-loopSkill
- freeze-reviewSkill
- governance-syncSkill
- print-nextSkill
- workspace-statusSkill
- build-verification-reviewSkill
- deploy-verification-reviewSkill
- freeze-review-loopSkill
- freeze-reviewSkill
- governance-syncSkill
- knowtationSkill
- print-nextSkill
- workspace-statusSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 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.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- could be automatic: a setting could enforce it
- Distinctive: line 12
Use when the operator says Check OK, check ok, Check if OK, or
- Distinctive: line 57
If the operator names a freeze/spec path → use it.
- Distinctive: line 72
Follow the freeze-review skill (or freeze-review-loop for rounds).
- Distinctive: line 83
Follow the build-verification-review skill.
- Very specific: line 84
Do not claim DONE / “shipped” without pass.
- Distinctive: line 91
When the session Model is Thinking (or Thinking → Auto):
- Very specific: line 104
Never open a new governance lane for a one-off check.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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
Check OK
Use when the operator says Check OK, check ok, Check if OK, or
/check-ok — for roadmap phases or side research that is not on the
ROADMAP. Does not open a new docs.lanes entry.
Purpose
Same twin honesty gates as roadmap re-review loops:
| Gate | When | Tool |
| Freeze review | Spec / Thinking output ready (frozen: true) | This skill → /freeze-review / /freeze-review-loop + ok review --freeze / ok check-ok |
| Build verification | Implementation claims done | /build-verification-review |
Engine is kit-owned (tools/freeze_reviewer/ + cli/commands/review.py) — identical in
Scooling, Knowtation, VideoFactory, and any repo after ok sync.
Where this skill is installed
| Runtime | Path after ok sync | How to invoke |
| Cursor | .cursor/skills/check-ok/ | Type Check OK or /check-ok |
| Claude Code | .claude/skills/check-ok/ | Type Check OK or /check-ok |
| Copilot / any chat | N/A (no native skills) | Paste docs/CHECK-OK.md or run CLI below |
| Any terminal | CLI in kit | ok check-ok … |
How the operator runs it
Chat (Cursor or Claude Code): type Check OK or invoke /check-ok.
Terminal (works everywhere — Copilot, ChatGPT paste sessions, no IDE skills):
ok check-ok --topic "my-side-research" --dry-run
ok check-ok --path docs/reviews/2026-07-17-my-side-research.md
ok check-ok --path docs/archive/phases/PHASE-FOO.md # existing freeze artifact(ok check-if-ok is a deprecated synonym for the same command.)
Chat does the semantic thinking-model review. Terminal runs the mechanical checklist
gate (same as ok review --freeze).
Workflow (execute in this session)
- Identify the artifact
- If the operator names a freeze/spec path → use it.
- Else if a PHASE / side-check doc already exists for the work → use it.
Else scaffold under
docs/reviews/<YYYY-MM-DD>-<slug>.mdvia:ok check-ok --topic "<short-slug>" --scaffold-onlyOr create the file by hand using the template in this skill directory (
SIDE-CHECK-TEMPLATE.md).
- Fill scope — intent, interfaces, fail-closed rules, seven-tier test matrix, what must not change. Commit on a feature branch (Tier 1).
- Freeze review (beginning gate) — Model:
thinking-high- Follow the freeze-review skill (or freeze-review-loop for rounds).
Run mechanical gate:
ok check-ok --path <artifact> --dry-run ok check-ok --path <artifact> # stamps on pass when configured(
ok review --freeze <artifact>is equivalent — same engine.)
- If Auto / code landed — build verification (end gate)
- Follow the build-verification-review skill.
- Do not claim DONE / “shipped” without
pass.
- Report — verdict
pass|findings|blocked; every finding needspath:line. Append the Review-record table on the artifact.
Thinking sessions (always-on)
When the session Model is Thinking (or Thinking → Auto):
- End of Thinking / before Auto: freeze review must reach
pass(step 3). - After Auto / implementation: build verification must reach
pass(step 4).
See the always-on rule check-ok-thinking.mdc (Cursor) and AGENTS.md § Check OK (all tools).
Hard stops
Escalate to human — do not greenwash — for:
security·irreversible·real_money·gates_tier3
Never open a new governance lane for a one-off check. Never merge to
{{vcs.git.main_branch}} as part of this skill (Tier 3).
Read first
docs/CHECK-OK.md— portable paste prompt (any assistant)docs/OVERSEER-KIT-SPEC.md§6- freeze-review + build-verification-review skills (same footprint)
.overseer/config.yaml→freeze_contract
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.

