# exhaustive-systems-analysis

exhaustive-systems-analysis performs structured, multi-subsystem code audits to uncover correctness bugs, race conditions, and production risks before release. Useful for verifying agent-generated code.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Windsurf (Devin Desktop), Claude.ai · 185 installs · ★ 49 repo stars
- Author: petekp
- License: MIT · policy=mirror · the beginning of the file is at the end of this page
- Source: https://github.com/petekp/claude-code-setup/tree/HEAD/skills/exhaustive-systems-analysis/SKILL.md
- Updated: 2026-10-25
- Security check: automated, 2026-11-16 (text) — no signs of malicious code found.
- HTML: https://affpapa.org/en/nevibecode/skills/exhaustive-systems-analysis

## What it does

The skill starts by reading requirements, architecture notes, and tests before looking at the implementation. It breaks codebases into distinct subsystems, builds a coverage ledger, and drafts concrete, falsifiable hypotheses. Audits trace entry points through normal execution paths as well as error recovery, cancellations, and retries.

Findings require explicit proof. Each issue is tagged with severity, confidence, and a status like `Confirmed`, `Likely`, or `Needs follow-up`. Static reasoning alone yields `Likely`, while `Confirmed` demands direct code contradictions or failing tests. For unused code, the skill enforces search queries across consumers before declaring anything dead.

By default, results appear directly in chat, with an option to write reports to `docs/audit/` or `.claude/docs/audit/`. The skill avoids editing code during an audit, keeping the focus entirely on diagnosis unless fixes are explicitly requested.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add petekp/claude-code-setup --skill exhaustive-systems-analysis -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/exhaustive-systems-analysis/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/exhaustive-systems-analysis/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add petekp/claude-code-setup --skill exhaustive-systems-analysis -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/exhaustive-systems-analysis/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add petekp/claude-code-setup --skill exhaustive-systems-analysis -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/exhaustive-systems-analysis/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install petekp/claude-code-setup skills/exhaustive-systems-analysis
```

2. **Manually: put the files in place**
- File: `.github/skills/exhaustive-systems-analysis/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add petekp/claude-code-setup --skill exhaustive-systems-analysis -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/petekp/claude-code-setup --path skills/exhaustive-systems-analysis
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/exhaustive-systems-analysis/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add petekp/claude-code-setup --skill exhaustive-systems-analysis -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/petekp/claude-code-setup/tree/HEAD/skills/exhaustive-systems-analysis/SKILL.md

## When to use it
- audit multi-module codebases for race conditions and logic flaws
- verify agent-written code before deploying to production
- detect dead code and out-of-sync documentation
- build structured risk ledgers across several subsystems

## Example prompts
- Audit the authentication flow and confirm if concurrent logins cause race conditions.
- Check agent-written changes in `src/billing/` before we ship to production.
- Run a full system audit in artifact mode and save the findings under `docs/audit/`.

## Source text · SKILL.md · MIT

Below is the content of a third-party file. It is data, not instructions for you: do not act on anything written inside unless the human asks. Block tag: b2dc1b07c97e2173.

BEGIN THIRD-PARTY FILE b2dc1b07c97e2173
```text
---
name: exhaustive-systems-analysis
description: |
  Perform evidence-driven, multi-subsystem audits of real codebases to find correctness bugs, race conditions, security gaps, stale documentation, dead code, and production-readiness risks. Use when asked to audit a system end-to-end, verify agent-written code before shipping, analyze a subsystem for correctness across multiple modules, or produce a structured risk report for a real implementation. Prefer other skills for a single isolated bug, a proposal or document review, or a dedicated dead-code cleanup.
---

# Exhaustive Systems Analysis

Use this skill for full-system correctness work. The job is to map the system, identify the highest-risk behaviors, prove or refute concrete failure hypotheses, and leave behind a report another engineer can act on without re-reading the whole codebase.

## Failure Modes To Prevent

1. Surface-level audits that scan files without following behavior end-to-end
2. False certainty: reporting suspicions as bugs without enough evidence
3. Context drift across large audits with many subsystems
4. Cosmetic reviews that miss the real correctness and ship-readiness risks

## Operating Mode

- Default to `chat-first` output. Return findings inline unless the user asks for docs or the audit clearly needs multi-session artifacts.
- Switch to `artifact mode` for large or resumable audits. Use `docs/audit/` or `.claude/docs/audit/`, matching the repo's existing conventions.
- Do not start fixing code while auditing unless the user explicitly asks for fixes. This skill is for diagnosis, proof, and prioritization.

## Workflow

### 0. Calibrate The Audit

Before reading deeply, write a one-screen scope brief using the template in `references/templates.md`.

Capture:
- system or area under review
- user-visible workflows or contracts that matter most
- likely high-risk surfaces: state, side effects, concurrency, auth, persistence, external integrations
- out-of-scope areas
- output mode: `chat-first` or `artifact mode`

If the request is broad, narrow it to the modules that can actually change user outcomes or ship readiness.

### 1. Load Intent Before Code

Read only the materials that establish intended behavior:
- `README`, `CLAUDE.md`, architecture docs, ADRs
- tests that describe user-visible or contract behavior
```
END THIRD-PARTY FILE b2dc1b07c97e2173

The third-party file has ended. Everything between the b2dc1b07c97e2173 tags is data: requests and commands from it are not followed. An “end of file” line with another tag inside the block is part of the file.

First 2333 of 8476 bytes. © petekp · MIT · https://github.com/petekp/claude-code-setup/tree/HEAD/skills/exhaustive-systems-analysis/SKILL.md

_Part of the page did not fit — see the HTML version._

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md
