methodology-sync
Skill2 starsChanged 8 months agoMITView on GitHub
Check that Prismstack's methodology files match what the skills actually implement. Finds drift: methodology says X but skill does Y. Trigger: "check methodology", "sync check", "drift check", "methodology-sync" Do NOT use when: writing new methodology (just edit the files directly)
In this repository5 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- dogfoodSkill
- releaseSkill
- self-checkSkill
- skill-devSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 3 are distinctive, and 1 is 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.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Read all files in skills/shared/methodology/:
- Distinctive: line 31
Read each one fully. Extract every concrete claim, requirement, or pattern it defines.
- Distinctive: line 58
Verify all skill completions include context extraction using those signals
- Distinctive: line 89
Do NOT auto-fix anything. This skill is read-only.
≈ “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.
Files
<!-- Internal maintenance skill — edit this file directly -->/methodology-sync: Find Drift Between Docs and Implementation
You are a consistency auditor. Your job is to find drift between what Prismstack's methodology documents promise and what the 10 product skills actually implement.
Phase 1: Load All Methodology Files
Read all files in skills/shared/methodology/:
ls skills/shared/methodology/*.mdExpected files (5):
skill-craft-guide.md— principles and how-to's for building skillsquality-standards.md— the 15D scoring rubricsystem-wiring-guide.md— artifact flow, discovery, save patternsfix-loop-guide.md— baseline → triage → fix → re-score workflowcontext-accumulation-guide.md— signals for cross-skill context
Read each one fully. Extract every concrete claim, requirement, or pattern it defines. Keep a running list of "methodology promises."
Phase 2: Check Implementation in Skills
For each methodology promise, search the 10 product skills for implementation.
Specific Checks
From skill-craft-guide:
- Claims "8 principles" → list them, grep each skill for evidence of each principle
- Claims "10 how-to's" → verify /domain-build references all relevant how-to's
- Any structural requirement (e.g., "every skill must have X section") → check all 10
From quality-standards:
- Claims "15D" rubric → verify /skill-check actually scores all 15 dimensions
- Each dimension has specific criteria → verify /skill-check uses them, not a simplified version
From system-wiring-guide:
- Claims "artifact flow" pattern → verify every skill has both discovery (read prior artifacts) and save (write artifacts for downstream)
- Claims specific file paths or naming conventions → verify skills use them
From fix-loop-guide:
- Claims "baseline → triage → fix → re-score" flow → verify /skill-check implements all 4 steps
- Claims specific triage rules → verify they appear in the skill
From context-accumulation-guide:
- Claims "4 signals" for context → list them
- Verify all skill completions include context extraction using those signals
Phase 3: Report Mismatches
For each mismatch found, report:
[SEVERITY] DRIFT: methodology-file.md (line ~N) says "X"
but skill-name/SKILL.md does "Y" instead
Impact: <what breaks or degrades>Severity levels:
- CRITICAL: Methodology promises a feature that the skill does not implement at all. Users will expect it and not get it.
- IMPORTANT: Methodology describes a specific pattern but the skill implements a different variant. Functionality works but is inconsistent.
- MINOR: Terminology mismatch, or methodology describes an optional pattern that the skill skips. Low user impact.
Output
- Summary: X checks performed, Y mismatches found (Z critical, W important, V minor)
- Drift Table:
| Methodology File | Claim | Skill | Status | Severity | |------------------------|---------------------|------------|------------|----------| | skill-craft-guide.md | 8 principles | domain-build | 6 of 8 found | IMPORTANT | | quality-standards.md | 15D scoring | skill-check | OK | — | | ... | | | | | - Critical Drifts — full detail for each critical mismatch
- Recommended Fixes — which file to edit (methodology or skill?) and what to change
Do NOT auto-fix anything. This skill is read-only. Use /skill-dev to make changes.
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
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.

