agentleFS
imMamdouhaboammarimMamdouhaboammar/dokion/skills/dokion-hardening/SKILL.md

dokion-hardening

Skill1 starsChanged 4 months agoMITView on GitHub

Runs and resumes a user-authored Dokion hardening playbook, records findings and evidence, validates repairs, and reports scoped readiness. Use when a repository contains .dokion/playbook.json or the user asks to run, resume, inspect, validate, or report a Dokion hardening loop.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 4 are distinctive, and 8 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 8

    Very specific

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

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

    Use the Dokion CLI as the execution boundary.

    line 8
  2. Very specific:

    Never select, install, substitute, reorder, or enable a capability.

    line 10
  3. Very specific:

    Read SPEC.md when the task concerns Dokion internals.

    line 14
  4. Very specific:

    Run dokion doctor and dokion validate.

    line 15
  5. Distinctive:

    If no active playbook exists, stop and explain that the user must create or copy one.

    line 16
  6. Very specific:

    Run dokion run only after validation succeeds.

    line 17
  7. Very specific:

    Read .dokion/state.json and HARDENING.md.

    line 21
  8. Distinctive:

    If the run is waiting for approval, present the exact subject and scope.

    line 23
  9. Very specific:

    After the user records a decision, run dokion resume.

    line 24
  10. Very specific:

    Do not infer approval from conversational tone when Dokion requires an approval record.

    line 34
  11. Distinctive:

    every declared verification command exits with code 0

    line 64
  12. Distinctive:

    Do not replace Bun with npm, yarn, or pnpm in this repository.

    line 80

≈ “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

Dokion Hardening

Use the Dokion CLI as the execution boundary. .dokion/playbook.json is the only execution authority.

Never select, install, substitute, reorder, or enable a capability. Never upgrade a capability or edit .dokion/playbook.json. Recommendations are inert until the user changes the playbook.

Start

  1. Read SPEC.md when the task concerns Dokion internals.
  2. Run dokion doctor and dokion validate.
  3. If no active playbook exists, stop and explain that the user must create or copy one.
  4. Run dokion run only after validation succeeds.

Resume

  1. Read .dokion/state.json and HARDENING.md.
  2. Run dokion status and dokion findings.
  3. If the run is waiting for approval, present the exact subject and scope.
  4. After the user records a decision, run dokion resume.

Approval commands

dokion approve finding:<id> --by <identity> --notes "<reason>"
dokion reject finding:<id> --by <identity> --notes "<reason>"
dokion approve step:<id> --by <identity> --notes "<reason>"

Do not infer approval from conversational tone when Dokion requires an approval record.

Loop Engineering commands

dokion loop audit
dokion loop init --pattern <daily-triage|test-driven-loop|bug-fix-loop>
dokion loop cost
dokion loop sync
dokion loop context

Memory Engineering commands

dokion memory audit
dokion memory init --pattern <session-scratchpad|project-episodic-log|durable-facts-store|retrieval-budget|memory-hygiene-loop>
dokion memory status
dokion memory patterns

Verification

A repair is accepted only when:

  1. the declared remediation command exits successfully
  2. the repair delta stays within the declared write scope
  3. suppression and test-deletion checks pass
  4. a required regression test was added or modified
  5. every declared verification command exits with code 0

A failed or rejected repair must be restored to its pre-remediation snapshot. Treat REPAIR_REJECTED, FAILED, BLOCKED, and TAINTED as non-ready states.

Development commands

Dokion itself uses Bun and TypeScript.

bun install --frozen-lockfile
bun test
bun run typecheck
bun run validate:contracts
bun run build

Do not replace Bun with npm, yarn, or pnpm in this repository.

Completion language

Report only what the configured playbook proved for the recorded commit. Include active degradations, skipped lanes, manual checks, and uncovered areas. Never state that a repository is generally production ready from a partial Dokion run.

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.