agentleFS
Sign inStart free
MrinankaMrinanka/aegiscontext/.claude/skills/plan-step/SKILL.md

plan-step

Skill1 starsChanged 11 days agoApache-2.0View on GitHub

Plan the next build step (use in plan mode), then build it after approval.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. kkokosa/dotLLM★ 528same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository3 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    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. Distinctive:

    No new dependency beyond the approved plan without asking me.

    line 29
  2. Distinctive:

    Never change or delete an existing test, or weaken a security rule, to get green.

    line 30
  3. Distinctive:

    If something surprises you (an API differs from the docs, a test fails for an unclear reason), stop and tell me instead of working around …

    line 31
  4. Very specific:

    When it's built, tell me to run /verify.

    line 34

≈ “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.

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

Find the first unticked step in docs/engineering/progress.md. That is this session's step. Never start another one.

Read: CLAUDE.md, docs/engineering/progress.md (including Known issues and Gotchas), the step's section in docs/engineering/plan.md, the spec sections it refers to, and the entries in docs/engineering/decision-log.md and docs/engineering/findings-step0.md that touch this step. Don't build on a [Step 0] item that findings haven't confirmed. Don't reopen an accepted decision without saying so.

Check current docs for every API this step uses: the Claude Code mod API types (the file the plugin-authoring skill names; grep for the event or noun), Node.js docs, and the pinned Gitleaks source. Tell me the exact versions you'll pin and the docs pages or file:line you read.

Give me the plan, no code yet:

  1. The step's goal in two sentences, and the spec § / R-ids it covers.
  2. Files to create or change, and what each will contain.
  3. Key functions and data shapes.
  4. Tests you'll write first, including the step's "Done when" check as an automated test where possible. Include failure cases: bad input, timeouts, CRLF, Unicode, huge input, a missing OS password store, a fake secret that must not reach a log.
  5. Assumptions the plan doesn't state.
  6. Anything in the step that looks wrong, risky or unclear, with your proposed fix.
  7. What is explicitly NOT part of this step.

Design choices: ask with AskUserQuestion, following "Questions" in CLAUDE.md. Then ask separately for plan approval; answers to design questions are not approval.

Step 0 only: the "build" is small probes in experiments/ that I run by hand (new sessions, VS Code, a VM). Give me numbered copy-paste steps, read the logs yourself afterwards, and record each check in docs/engineering/findings-step0.md: result, claude --version, OS, surface, method, short evidence excerpt (no secrets), spec impact.

When I approve, build with these rules:

  • Tests first, and show they fail for the right reason. Then the code.
  • Small increments; run npm test after each one.
  • No new dependency beyond the approved plan without asking me.
  • Never change or delete an existing test, or weaken a security rule, to get green. If a test seems wrong, stop and explain.
  • If something surprises you (an API differs from the docs, a test fails for an unclear reason), stop and tell me instead of working around it.
  • Anything deferred goes to spec §23 (a feature) or progress "Known issues" (a bug or unfinished bit).

When it's built, tell me to run /verify.

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.

Check your setup, free