agentleFS
Sign inStart free
fagemxfagemx/prismstack/.claude/skills/self-check/SKILL.md

self-check

Skill2 starsChanged 8 months agoMITView on GitHub

Audit Prismstack's own skills using its own 15D quality rubric. Trigger: "check our skills", "self audit", "eat our own dog food", "self-check" Do NOT use when: checking a user's domain stack (use the product /skill-check for that)

11 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 8 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read skills/shared/methodology/quality-standards.md from the Prismstack project root.

    line 17
  2. Distinctive:

    If the file is missing or unreadable, stop and report: "Cannot run self-check — quality-standards.md not found."

    line 21
  3. Distinctive:

    After scoring all 10, look for systemic patterns:

    line 49
  4. Very specific:

    Check if .claude/skills/self-check/last-results.json exists.

    line 58
  5. Distinctive:

    If it does, load it and compute deltas:

    line 60
  6. Distinctive:

    If no previous results exist, note: "First run — no baseline for comparison."

    line 65
  7. Very specific:

    Write results to .claude/skills/self-check/last-results.json:

    line 78

“Common” is measured: kept by at least 6 in 10 of the 8 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
<!-- Internal maintenance skill — edit this file directly -->

/self-check: Audit Prismstack With Its Own Rubric

You are Prismstack's internal quality auditor. Your job is to score Prismstack's own 10 product skills using the same 15D rubric we ship to users. Eating our own dog food.

Phase 1: Load the Rubric

Read skills/shared/methodology/quality-standards.md from the Prismstack project root.

Extract all 15 dimensions and their scoring criteria. These are the exact dimensions you will score against — do not invent your own.

If the file is missing or unreadable, stop and report: "Cannot run self-check — quality-standards.md not found."

Phase 2: Discover All Product Skills

ls skills/*/SKILL.md

Expect 10 product skills. Read each SKILL.md fully. If fewer than 10 are found, note which are missing.

Phase 3: Score Each Skill on 15D

For each skill, score every dimension 1-5 using the rubric criteria from Phase 1.

Output a summary table:

| Skill           | D1 | D2 | D3 | ... | D15 | Total | Avg  |
|-----------------|----|----|----|----- |-----|-------|------|
| domain-plan     |  4 |  5 |  3 | ... |   4 |   58  | 3.87 |
| domain-build    |  3 |  4 |  4 | ... |   3 |   52  | 3.47 |
| ...             |    |    |    |     |     |       |      |

For any score below 3, add a one-line explanation of what is lacking.

Phase 4: Cross-Skill Pattern Analysis

After scoring all 10, look for systemic patterns:

  • Weak dimensions: Any dimension averaging below 3.0 across all skills? This is a methodology problem, not a skill problem.
  • Strong dimensions: Any dimension averaging above 4.5? Document what we are doing right.
  • Outlier skills: Any skill more than 1.0 below the average total? Flag for priority improvement.
  • Consistency: Are similar skills (e.g., domain-plan vs domain-build) scored similarly on shared dimensions?

Phase 5: Compare to Last Run

Check if .claude/skills/self-check/last-results.json exists.

If it does, load it and compute deltas:

  • Per-skill total change (improved / regressed / unchanged)
  • Per-dimension average change
  • Highlight any dimension that dropped by more than 0.5

If no previous results exist, note: "First run — no baseline for comparison."

Output

Present in this order:

  1. Score Table (Phase 3)
  2. Systemic Issues (Phase 4) — bullet list with severity
  3. Delta Report (Phase 5) — improvements and regressions since last run
  4. Top 3 Recommendations — the highest-impact fixes, with which skill and dimension to target

Save Results

Write results to .claude/skills/self-check/last-results.json:

{
  "timestamp": "YYYY-MM-DDTHH:MM:SS",
  "version": "<from VERSION file>",
  "skills": {
    "domain-plan": { "scores": { "D1": 4, "D2": 5, ... }, "total": 58, "avg": 3.87 },
    ...
  },
  "dimension_averages": { "D1": 3.8, "D2": 4.1, ... },
  "overall_average": 3.72
}

Confirm the file was saved so the next run can compare.

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