agentleFS
shapshap/shap/.claude/skills/ai-disclosure/SKILL.md

ai-disclosure

Skill26k starsChanged 9 months agoMITView on GitHub

Track and document Claude's contributions during coding sessions for PR transparency. Use when working on feature branches, PRs, or when the user wants to maintain AI contribution records. Maintains a disclosure file per branch summarizing Claude's involvement.

3 other projects use a version of this.

Adopted by

Steady: 0 new projects took it in the last 4 weeks, none the 4 weeks before.thin data

Week of Sep 20: 0 new projectsWeek of Sep 27: 0 new projectsWeek of Oct 4: 0 new projectsWeek of Oct 11: 0 new projectsWeek of Oct 18: 0 new projectsWeek of Oct 25: 0 new projectsWeek of Nov 1: 0 new projectsWeek of Nov 8: 0 new projects
Sep 2026Oct 2026Nov 2026
New projects taking a version of this file each week, over the 8 weeks our latest run read: only 0 in all, so this is a hint, not a trend.

Most recently taken up by

Who's talking about shap/shap

The biggest 5 of 65 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 65

Talked about on GitHub, YouTube, Hacker News and Stack Overflow (65 times).

On GitHub · 50

45 more on GitHub

On YouTube · 13

8 more on YouTube

On Hacker News · 1

  • Other versions3 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 3

    1. pradyumnasagar/open-research-skills★ 8same name
    2. nguyenbui1105/first-1000-players★ 1same name
    3. Ricardo-kui/claude-skills★ 0same 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.

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 13 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈13

    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:

    When starting work on a feature branch, offer once:

    line 26
  2. Distinctive:

    If yes, create the disclosure file and begin tracking.

    line 30
  3. Distinctive:

    Only track contributions made AFTER the user opts in: Do not backfill anything from earlier in the conversation, even if it seems relevant …

    line 34
  4. Distinctive:

    After significant actions (writing functions, fixing bugs, refactoring):

    line 36
  5. Distinctive:

    Only use the disclosure file to track contributions — do not rely on in-context memory of what happened before opt-in.

    line 44
  6. Distinctive:

    No prompts during work—just silently maintain the record.

    line 46
  7. Distinctive:

    When user requests summary or creates a PR:

    line 50
  8. Distinctive:

    Use git diff or read the files to see if user modified them after you

    line 55
  9. Distinctive:

    If user significantly modified your code afterward → downgrade to Assisted

    line 59
  10. Distinctive:

    If user corrected your approach multiple times → downgrade to Assisted

    line 60
  11. Distinctive:

    Add note: "co-creation with significant user involvement"

    line 61
  12. Distinctive:

    Downgrade rule: If user significantly modified your code or corrected your approach repeatedly, downgrade from Autonomous to Assisted and …

    line 103
  13. Distinctive:

    When user creates a PR, offer a copy-paste block:

    line 127

≈ “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

AI Disclosure Tracking

Track Claude's contributions during coding sessions and maintain a disclosure file for PR transparency.

Involvement Levels

  1. Autonomous - Claude wrote the code/solution independently
  2. Assisted - Claude implemented based on user direction
  3. Advised - Claude provided guidance that user implemented

File Location

Disclosure file: .claude/disclosures/<branch-name>.md

Get branch name with git branch --show-current. If not in a git repo, use session-<date>.

Workflow

1. Opt-in

When starting work on a feature branch, offer once:

"Would you like me to track AI contributions for this branch?"

If yes, create the disclosure file and begin tracking.

2. Track Silently

Only track contributions made AFTER the user opts in. Do not backfill anything from earlier in the conversation, even if it seems relevant to the branch. The disclosure file starts as a clean slate from the moment the user says yes.

After significant actions (writing functions, fixing bugs, refactoring):

  1. Log the contribution - Append to the appropriate section in the disclosure file
  2. Record what you changed - Track in the internal section:
    • Which files you touched
    • Brief summary of what you wrote/changed
    • Initial involvement level

Only use the disclosure file to track contributions — do not rely on in-context memory of what happened before opt-in.

No prompts during work—just silently maintain the record.

3. Verify and Generate Summary

When user requests summary or creates a PR:

  1. Check your work against current state:
    • Review files you logged as touching
    • Compare current file contents against what you originally wrote
    • Use git diff or read the files to see if user modified them after you
    • Also recall from conversation context: did user correct you? Ask for changes? Rewrite parts?
  2. Downgrade if needed:
    • If user significantly modified your code afterward → downgrade to Assisted
    • If user corrected your approach multiple times → downgrade to Assisted
    • Add note: "co-creation with significant user involvement"
  3. Generate the summary with accurate involvement levels

Disclosure File Format

# AI Disclosure for branch: <branch-name>

## Summary
[Generated on request]

## Contributions

### Autonomous
- [One-line descriptions of independent work]

### Assisted
- [One-line descriptions of directed work]

### Advised
- [One-line descriptions of guidance provided]

Internal Tracking

Track your changes in an HTML comment (not shown in final summary):

<!--
CHANGES:
- src/partition.py: wrote repartitioning logic (autonomous)
- tests/test_partition.py: wrote validation tests (autonomous)
- src/boundaries.py: implemented boundary calc (autonomous)

CORRECTIONS:
- src/boundaries.py: user fixed off-by-one error (count: 2)
-->

This record lets you verify at summary time whether files still contain what you wrote, or if the user significantly changed them.

Downgrade rule: If user significantly modified your code or corrected your approach repeatedly, downgrade from Autonomous to Assisted and note: "co-creation with significant user involvement".

Example Output

# AI Disclosure for branch: feature/healpix-partitioning

## Summary
Claude assisted with repartitioning logic (co-creation with significant human involvement), autonomously wrote test cases, and advised on spatial indexing approaches.

## Contributions

### Autonomous
- Wrote test cases for HEALPix partition validation

### Assisted
- Implemented repartitioning logic based on user requirements

### Advised
- Suggested using spatial indexing for performance

PR Format

When user creates a PR, offer a copy-paste block:

## AI Disclosure

Developed with Claude assistance:
- **Autonomous**: [list]
- **Assisted**: [list]
- **Advised**: [list]

Details: `.claude/disclosures/<branch>.md`

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.