agentleFS
Sign inStart free
sson0-ersson0-er/claudex/skills/record-evidence/SKILL.md

record-evidence

Skill0 starsChanged 53 days agoMITView on GitHub

How to record external facts (docs, command output, codebase findings) as evidence and design choices as decisions. Used by the researcher, designer and reviewer agents.

We split this skill into the rules it gives an agent

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

  • 4

    Common

    Most of the 3 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Common:

    Write one evidence entry immediately after any of these:

    line 11
  2. Common:

    Write one decision entry whenever you choose between alternatives (adopting an approach, rejecting another).

    line 17
  3. Very specific:

    Create docs/claudex/evidence/<yyyy-mm-dd>-<slug>.md using the format in ${CLAUDE_PLUGIN_ROOT}/templates/evidence.md.

    line 21
  4. Very specific:

    Create docs/claudex/decisions/<nnn>-<slug>.md using ${CLAUDE_PLUGIN_ROOT}/templates/decision.md.

    line 31
  5. Very specific:

    When a design, research note or review finding states an external fact, end the sentence with (evidence: <id>).

    line 36
  6. Common:

    If you cannot cite, write "unverified".

    line 37
  7. Very specific:

    Before relying on an entry past its expires date, re-check the source and update fetched and expires.

    line 41
  8. Common:

    If you cannot re-check, state "expired, not re-verified" where you use it.

    line 42

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “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

Recording evidence and decisions

When to record

Write one evidence entry immediately after any of these:

  • You consulted an external source: official docs, source code, an issue, an RFC.
  • You ran a command to confirm behavior or a version (CLI flags, API responses, installed versions).
  • You read existing code and found a fact that a design decision will rest on.

Write one decision entry whenever you choose between alternatives (adopting an approach, rejecting another).

Writing an evidence entry

  1. Create docs/claudex/evidence/<yyyy-mm-dd>-<slug>.md using the format in ${CLAUDE_PLUGIN_ROOT}/templates/evidence.md. The slug uses [a-z0-9-] only.
  2. Body is a summary: a few lines up to a dozen. Never paste whole pages or raw command output.
  3. fetched is today's date. expires is 90 days later by default, 30 days for fast-moving sources (preview features, beta APIs).
  4. Separate what you learned from what you could not confirm.
  5. Exclude secrets: tokens, output of authenticated commands, internal URLs, personal data.
  6. Append one line to docs/claudex/evidence/INDEX.md: - <id> | <question> | expires <yyyy-mm-dd>

Writing a decision entry

  1. Number it <nnn> (three digits, zero-padded) by counting existing files in docs/claudex/decisions/.
  2. Create docs/claudex/decisions/<nnn>-<slug>.md using ${CLAUDE_PLUGIN_ROOT}/templates/decision.md.
  3. In "Rationale", reference evidence ids. If a reason has no evidence, write "unverified" next to it.

Citing in deliverables

When a design, research note or review finding states an external fact, end the sentence with (evidence: <id>). If you cannot cite, write "unverified". Reviewers treat uncited external claims as findings.

Expired evidence

Before relying on an entry past its expires date, re-check the source and update fetched and expires. If you cannot re-check, state "expired, not re-verified" where you use it.

Download now (ZIP)

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 2

    1. sson0-er/orchestraude★ 092% same
    2. sson0-er/sonnede★ 094% same

    Wrote their own version · 1

    1. MarkTripoli/skills★ 1same 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 repository2 other files it gives its agents.

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