agentleFS
Wsh7AshWsh7Ash/skillmcp/skills/description-as-contract/SKILL.md

description-as-contract

Skill0 starsChanged 15 days agoMITView on GitHub

Write tool descriptions as the contract a model will follow, including what the tool does not do. Use when an agent must choose between tools without seeing the implementation

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Keep the description free of implementation detail the model cannot act on

    line 23
  2. Distinctive:

    Put constraints in the same order the model must respect them

    line 24
  3. Distinctive:

    no description relies on source-reading to be understood

    line 63

≈ “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

Description as Contract

Write tool descriptions as the contract a model will follow, including what the tool does not do

Use when: an agent must choose between tools without seeing the implementation

Avoid when: a human is the only caller and reads the source directly

Procedure

  1. State what the tool returns in one sentence
  2. State what it does not do, such as no network access or no writes
  3. Name the failure modes a caller will encounter
  4. Keep the description free of implementation detail the model cannot act on
  5. Put constraints in the same order the model must respect them
  6. Re-read the description as if it were the only information available

Signals

  • the description names the return shape and its bounds
  • a caller knows which sibling tool to use next
  • limitations appear in the description, not only in an error
  • the text has no marketing adjectives

Patterns

1. Return statement

ranks skills by keyword over slug, title, tags and description

2. Negation

does not read outside the packs root and sends no network traffic

3. Failure note

returns isError with a suggestion when nothing matches

Pitfalls

  • Describing the mechanism instead of the contract
  • Omitting limits so a model assumes completeness
  • Repeating the tool name in the description instead of adding constraints

Done when

  • each description states return shape, bounds and limitations
  • no description relies on source-reading to be understood

References

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.