agentleFS
youcislayoucisla/Agent-Foundry/skills/core/plan-then-act/SKILL.md

plan-then-act

Skill1 starsChanged 23 days agoMITView on GitHub

Plan first, then act. State the plan in one sentence before making the

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not narrate every tool call.

    line 28

≈ “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

Plan Then Act

State the plan once, execute it in bounded steps, and verify each step. This skill owns execution cadence; plan-before-code owns the spec gate.

Procedure

1 - State the plan

One sentence: goal, ordered phases, and the first action.

Plan: build [X] by [A], then [B], then verify with [C]. First: [action].

Do not narrate every tool call. A short plan prevents drift; constant commentary adds noise.

2 - Execute one bounded step at a time

Each step should produce an observable result. After a change, run the narrowest check that can fail: a targeted test, a typecheck, a request, a screenshot.

3 - Batch independent inspections

Group reads and searches that do not depend on each other. Serialize changes that touch the same file or depend on a previous result.

4 - Keep coherent code together

Co-locate tightly coupled logic in one module when that makes the behavior easier to follow. Split when responsibilities are independently testable, several consumers need only part of the module, or the file becomes hard to navigate. File length alone is not a reason to split.

5 - Verify behavior, not just compilation

For UI work, exercise the real page and inspect the rendered result. For API work, call the endpoint. For data changes, query the row. A successful build is not a behavior check.

6 - Stop on the defined condition

Name the stop condition in the plan: tests pass, the requested behavior is observed, or a checkpoint is reached. Do not keep expanding scope after it.

When to Use

  • Multi-step tasks where order matters
  • Multi-file changes
  • UI work where visual correctness matters
  • Work that can drift from the stated goal

When Not to Use

  • A single known-command execution
  • A pure read-only query

Anti-patterns

  • A plan that is restated after every tool call.
  • Writing files and moving on without a check.
  • Batching dependent calls and guessing the result.
  • Continuing past the stop condition into unrequested work.
  • Claiming success from a compile or a diff alone.

Verification Checklist

  • The plan and first action were stated before execution.
  • Steps ran in the planned order, or a deviation was explained.
  • Each change was followed by a targeted check.
  • The final check exercises the requested behavior.
  • The stop condition was met without scope creep.

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.