agentleFS
botdigit-officialbotdigit-official/agent-blueprint/skills/06-codebase-audit/SKILL.md

06-codebase-audit

Skill1 starsChanged 53 days agoMITView on GitHub

Version: 1.0.0 Compatible: project_skills >= 1.0 Requires: discovery, project-context Outputs: codebase-audit.md, code-gaps.md Inspect the existing codebase to understand what is there, how it is structured, what conventions it follows, and what problems exist — before making any changes. Inspect before assuming. For each major feature or flow: - Where it enters the system - What components it touches - What data it reads and writes - What external services it calls - What errors can occur Look for: - Duplicated…

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are 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.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Test quality (unit, integration, end-to-end)

    line 68
  2. Distinctive:

    Test coverage of critical paths

    line 69
  3. Distinctive:

    Test debt (missing tests, brittle tests, no CI)

    line 77

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 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

06-Codebase-Audit Skill

Version: 1.0.0 Compatible: project_skills >= 1.0 Requires: discovery, project-context Outputs: codebase-audit.md, code-gaps.md


Purpose

Inspect the existing codebase to understand what is there, how it is structured, what conventions it follows, and what problems exist — before making any changes.

Inspect before assuming.


Process

1. Map the code structure

  • Directory layout
  • Entry points
  • Module/package boundaries
  • Public vs private interfaces
  • Shared code
  • Standalone scripts

2. Identify conventions

  • Naming conventions (files, functions, variables, classes)
  • Code style (formatting, linting, imports)
  • Architectural patterns in use (MVC, layered, domain-driven, etc.)
  • Error handling patterns
  • Logging patterns
  • Testing patterns (if tests exist)
  • Configuration patterns

3. Identify entry points and flows

For each major feature or flow:

  • Where it enters the system
  • What components it touches
  • What data it reads and writes
  • What external services it calls
  • What errors can occur

4. Assess code quality

Look for:

  • Duplicated code
  • Large functions or files
  • Mixed responsibilities
  • Copy-paste architecture
  • Dead code
  • commented-out code
  • TODOs and FIXMEs (and whether they are tracked)
  • Magic values
  • Inconsistent error handling
  • Missing input validation
  • Missing type safety where the language supports it

5. Assess test situation

  • Do tests exist?
  • What is tested?
  • What is not tested?
  • Test quality (unit, integration, end-to-end)
  • Test coverage of critical paths
  • Whether tests are run in CI

6. Identify technical debt

Record debt in these categories:

  • Code-level debt (duplication, complexity, poor naming)
  • Architecture-level debt (wrong boundaries, coupling, missing abstractions)
  • Test debt (missing tests, brittle tests, no CI)
  • Documentation debt (missing, outdated)
  • Dependency debt (outdated, unmaintained, too many)
  • Security debt (see 07-security for detailed security audit)

Output

codebase-audit.md

Codebase Audit
==============
Project: [name]
Audit date: [date]
Auditor: [agent or person]

Structure:
  [description of code layout]

Conventions:
  [naming, style, patterns detected]

Entry points:
  [list]

Code quality:
  Strengths: [what is good]
  Problems: [what is not good]
  Estimated complexity: [low | medium | high]

Test situation:
  Tests exist: [yes | no]
  Coverage: [description or "unknown"]
  CI: [yes | no | unknown]

Technical debt:
  [list by category]

code-gaps.md

Code Gaps
=========
[Gap ID]: [description]
Severity: [low | medium | high]
Category: [code | architecture | test | docs | dependency]
Recommendation: [what to do]

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.

Reviews can't be read right now.

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.