agentleFS
gootfgootf/software-agent-sop/architecture-audit/SKILL.md

architecture-audit

Skill2 starsChanged 2 months agoMITView on GitHub

Comprehensive project consistency review across code, documentation, diagrams, and configuration

16 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

Used with

Projects with it also use prototype, review and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • GoogleSkill

    prototype

    Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

    Also in 3 projects that use this

  • SentrySkill

    review

    Fixture review skill.

    Also in 3 projects that use this

  • GoogleSkill

    tdd

    Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 11 edited versions we read keep these.

  • 2

    Distinctive

    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
  1. Distinctive:

    Before major releases or after significant refactoring

    line 29
  2. Very specific:

    Check: All version references match the canonical version (usually package.json)

    line 53
  3. Distinctive:

    Check: All documentation referencing these counts matches actual values

    line 89

“Common” is measured: kept by at least 6 in 10 of the 11 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

Architecture Audit

Comprehensive project consistency review across code, documentation, diagrams, and configuration

Overview

Systematic audit process to ensure all project artifacts stay synchronized. Catches version drift, terminology inconsistencies, outdated diagrams, broken references, and code-to-docs mismatches.

⚠️ IMPORTANT: This skill audits the user's project code, NOT the the AI assistant cognitive architecture in .github/. Ignore .github/ folder contents when performing audits - focus on the actual source code, documentation, and configuration in the project root and subdirectories.

Triggers

  • "audit", "comprehensive review", "fact-check"
  • "consistency check", "project health"
  • "pre-release audit", "documentation review"
  • Before major releases or after significant refactoring

Audit Checklist

1. Version Consistency

# Find version references in common locations
# EXCLUDE: .github/** (the AI assistant cognitive architecture - not project code)
$patterns = @(
    'package.json',           # "version": "x.y.z"
    'src/**/config*.json',    # Version in config files (not .github)
    '*.md',                   # Root documentation only
    'docs/**/*.md',           # Project docs (not .github)
    'src/**/constants.ts',    # Hardcoded versions
    'CHANGELOG.md'            # Version headers
)

# Grep for version patterns
Get-ChildItem -Recurse -Include $patterns |
    Select-String -Pattern 'v?\d+\.\d+\.\d+' |
    Group-Object -Property Line

Check: All version references match the canonical version (usually package.json)

2. Terminology Consistency

Build a deprecated terms list for your project:

Deprecated TermCurrent TermMigration Pattern
DK-*.mdskills/*/SKILL.mdFile format change
domain-knowledge/skills/Folder rename
(project-specific)(project-specific)(document here)
# Search for deprecated terms (exclude .github/)
$deprecated = @('OLD_TERM_1', 'OLD_TERM_2')
foreach ($term in $deprecated) {
    Get-ChildItem -Recurse -Include "*.md","*.ts","*.json" -Exclude ".github" |
        Where-Object { $_.FullName -notmatch '\\.github\\' } |
        Select-String -Pattern $term
}

3. Fact Inventory

Maintain a facts file or section with current counts:

## Current Facts (Update on Audit)
- Skills: 46 folders
- Commands: 16 total
- Tools: 11 MCP tools
- Agents: 5 custom agents
- Instructions: 19 files
- Prompts: 7 files

Check: All documentation referencing these counts matches actual values

4. Diagram Validation

For each diagram (Mermaid/ASCII):

  • Node labels match current terminology
  • Counts in labels are accurate
  • Flow directions reflect actual data flow
  • Subgroup names are current
  • No references to deprecated concepts
# Find all diagrams in project (exclude .github/)
Get-ChildItem -Recurse -Include "*.md" |
    Where-Object { $_.FullName -notmatch '\\.github\\' } |
    Select-String -Pattern '```mermaid|```ascii|┌|╔' |
    Group-Object -Property Path

5. Cross-Reference Validation

  • Internal links resolve ([text](path) → file exists)
  • Anchor links work (#section-name → heading exists)
  • Image references valid
  • Import/require statements resolve
# Find markdown links in project (exclude .github/)
Get-ChildItem -Recurse -Include "*.md" |
    Where-Object { $_.FullName -notmatch '\\.github\\' } |
    Select-String -Pattern '\[.*?\]\((?!http)[^)]+\)' |
    ForEach-Object {
        # Extract path and verify existence
    }

6. Code-to-Docs Sync

Code ArtifactDocumentation LocationCheck
Exported functionsAPI docsSignatures match
Config optionsREADME/User ManualAll options documented
CLI commandsHelp text / docsCommands listed
Error messagesTroubleshootingErrors explained

7. Configuration Alignment

  • TypeScript constants match documentation
  • JSON schemas match actual structure
  • Environment variables documented
  • Feature flags documented

Audit Report Template

# Architecture Audit Report
**Date:** YYYY-MM-DD
**Version Audited:** x.y.z
**Auditor:** [name/AI]

## Summary
| Category | Status | Issues Found |
|----------|--------|--------------|
| Version Consistency | ✅/⚠️/❌ | N |
| Terminology | ✅/⚠️/❌ | N |
| Fact Accuracy | ✅/⚠️/❌ | N |
| Diagrams | ✅/⚠️/❌ | N |
| Cross-References | ✅/⚠️/❌ | N |
| Code-Docs Sync | ✅/⚠️/❌ | N |

## Issues Found
### [Category]
- **File:** path/to/file
- **Line:** N
- **Issue:** Description
- **Fix:** Suggested correction

## Actions Taken
- [ ] Issue 1 fixed
- [ ] Issue 2 fixed

Automation Opportunities

Pre-commit Hook

# .husky/pre-commit or similar
npm run audit:quick  # Fast checks only

CI Pipeline

# GitHub Actions
- name: Architecture Audit
  run: npm run audit:full
  if: github.event_name == 'pull_request'

Scheduled Audit

# Weekly full audit
on:
  schedule:
    - cron: '0 9 * * 1'  # Monday 9am

Anti-Patterns

❌ Skipping audits before release — Drift accumulates silently

❌ Manual-only audits — Automate what you can

❌ Fixing without documenting — Record the pattern for next time

❌ Ignoring "cosmetic" issues — They signal deeper drift

❌ Auditing only changed files — Drift affects unchanged files too

Best Practices

✅ Run full audit before major releases

✅ Add new deprecated terms as you migrate

✅ Update fact inventory after structural changes

✅ Document audit findings for patterns

✅ Automate repetitive checks

✅ Time-box manual review (don't boil the ocean)

Integration with Other Skills

  • release-preflight → Trigger audit before release
  • brain-qa → Connection validation and architecture health
  • code-review → Audit as part of PR review
  • refactoring-patterns → Audit after major refactoring

your AI assistant Extensions

your AI assistant-specific audit procedures that extend the generic project audit above. These leverage knowledge of exact folder structure, heir relationships, safety imperatives, and build/release workflows.

Audit Philosophy (Semantic + Structural)

DimensionScript (Automated)the AI assistant (With This Skill)
StructuralFile existence, counts, format✅ Automated
Semantic—Do files describe concepts consistently?
Logic—Are documented workflows achievable?
CodeCompile, lint, bundleDoes code match documented claims?
ArchitecturalVersion alignment, heir syncIs memory architecture reflected accurately?

Master-Only Sections (1-9)

#SectionWhat It Checks
1Version Alignmentpackage.json ↔ CHANGELOG ↔ copilot-instructions
2Heir Folder SyncMaster → heir file counts
3Skill InheritanceSKILL_EXCLUSIONS in sync-architecture.cjs
4Safety ImperativesKill switch, protection markers
5Build Artifactsdist/extension.js exists and recent
6Documentation Cross-RefsRequired links between docs
7Connection HealthValid links, no orphans
8alex_docs AuditVersion refs, deprecated terms
9Skill Network DiagramNode count matches actual skills

Inheritable Sections (10-22)

#SectionWhat It Checks
10Extension UIDead buttons
11Dependency Healthnpm audit vulnerabilities
12TypeScript & LintCompile errors, lint warnings
13SecuritySecrets in code, CSP compliance
14Bundle Sizeextension.js size
15Git HygieneUncommitted changes, branch status
16ChangelogFormat, version alignment
17API CompatibilityDeprecated VS Code APIs
18Test CoverageTest file count vs source files
19AccessibilityARIA labels, CSS variables
20Localizationl10n configuration
21Asset IntegrityIcon exists, no missing assets
22Configuration Fileslaunch.json, tsconfig.json, .vscodeignore

Production Scripts

# Full audit (all 22 sections)
node .github/muscles/brain-qa.cjs

# Quick audit (sections 1-9)
node .github/muscles/brain-qa.cjs --mode quick

# Specific sections
node .github/muscles/brain-qa.cjs --mode quick --quiet

Semantic Review Checklist (Manual — Not Scriptable)

  • Cross-heir semantic parity: M365 system prompt matches copilot-instructions
  • Deprecated terminology: No "DK files" / "domain-knowledge/" in active files
  • Claim vs. reality: README features match package.json commands
  • Process logic: Heir evolution, meditation, dream workflows internally consistent
  • Architectural model: Memory architecture described consistently everywhere
  • Code behavior: TypeScript handlers match instruction file docs
  • Version source of truth: package.json is canonical

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.