Skip to main content
skill CX Codex

overnight-repo-auditor — Claude Code skill: how to install and use

overnight-repo-auditor runs an autonomous overnight code audit covering security, performance, accessibility, and dependencies without prompting for confirmation.

MIT 253 installs ★ 440 updated 2026-10-20

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add OneWave-AI/claude-skills --skill overnight-repo-auditor -a claude-code
Open on GitHub

What overnight-repo-auditor does

Built for task horizons up to 14.5 hours under Anthropic Managed Agents, this skill organizes repository reviews into three distinct phases. An initial reconnaissance pass scans project structure and language mix, logging context into audit-workspace/00-reconnaissance.md. It then triggers background sub-agents simultaneously for security, code quality, performance, UI accessibility, and manifest dependencies.

Operations remain read-only against your source tree throughout the entire run. Agents never rewrite project files; the only commands executed are read-only vulnerability checks such as npm audit or pip audit. Findings stream straight into audit-workspace/ to protect partial progress against interruptions before compiling into overnight-audit-report.md with an executive top-10 list. The entire setup requires the Agent tool to dispatch these concurrent passes.

What it helps with

  • audit an entire codebase overnight without interactive prompts
  • identify package vulnerabilities via npm audit and pip audit
  • detect UI accessibility issues and performance bottlenecks
  • compile a single severity report with top-10 remediation priorities

Example prompts for the agent

  • Run an overnight audit of the entire repository
  • Audit only security and dependencies for this codebase
  • Analyze project performance and code quality, then write the final report

How to install overnight-repo-auditor

1Skills CLI (Vercel)

npx skills add OneWave-AI/claude-skills --skill overnight-repo-auditor -a claude-code

2Manually: put the files in place

~/.claude/skills/overnight-repo-auditor/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/overnight-repo-auditor/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about overnight-repo-auditor

How to install overnight-repo-auditor in Claude Code?

Run npx skills add OneWave-AI/claude-skills --skill overnight-repo-auditor -a claude-code in your terminal.

What does overnight-repo-auditor do?

It scans repositories overnight across security, performance, dependencies, and code quality, compiling findings into overnight-audit-report.md.

Does overnight-repo-auditor modify source code?

No, it operates in read-only mode against project files, only writing to the audit-workspace/ folder and the final report file.

Source text

SKILL.md · MIT
---
name: overnight-repo-auditor
description: Uses Managed Agents' 14.5-hour runtime to audit an entire codebase overnight. Security, performance, accessibility, dependency issues. You wake up to a full report.
tools: Read, Grep, Glob, Bash, Agent, Write
model: inherit
---

# Overnight Repo Auditor

Autonomously audit an entire codebase overnight and produce a single severity-rated report covering security, performance, accessibility, dependencies, and code quality. Built for Anthropic's Managed Agents runtime (14.5-hour task horizon): run to completion without questions or confirmation, writing structured findings to disk as execution proceeds so partial results survive interruption.

## Contents

- `references/reconnaissance.md` -- Phase 1 steps and the reconnaissance report template.
- `references/shared-rubric.md` -- Severity rating rubric and structured finding format. Pass both to every agent.
- `references/agent-security.md` -- Security Auditor brief and output format.
- `references/agent-performance.md` -- Performance Auditor brief and output format.
- `references/agent-accessibility.md` -- Accessibility Auditor brief, skip condition, and output format.
- `references/agent-dependency.md` -- Dependency Auditor brief, skip condition, and output format.
- `references/agent-code-quality.md` -- Code Quality Auditor brief and output format.
- `references/compilation-and-templates.md` -- Phase 3 compilation steps, final report template, and completion message.

## Operating Rules

- Run autonomously. Never ask the user for input. When a decision is ambiguous, choose the more thorough option and document the choice in the report.
- Read-only against source. Create only the `audit-workspace/` directory and `overnight-audit-report.md`. Never modify, build, or execute project code. The sole exception: the Dependency Auditor may run read-only package-audit commands (`npm audit`, `pip audit`, and equivalents).
- Be exhaustive, not sampling-based, while the time window allows.

## Workflow

### Phase 1: Reconnaissance (sequential, ~5-10 min)

1. Scan repository structure, identify languages, frameworks, config files, and estimate lines of code.
2. Determine which audit modules are relevant (Security and Code Quality always; Performance always; Accessibility only with frontend/template files; Dependency only with a manifest/lockfile).
3. Write `audit-workspace/00-reconnaissance.md` as the shared context document for all agents.

Follow `references/reconnaissance.md` for exact commands and the report template.

### Phase 2: Parallel Audit Deployment

4. Deploy every relevant audit agent simultaneously via the Agent tool. Use `run_in_background: true` on every call and send ALL agent calls in a single message.
5. Build each agent's prompt by combining: the full reconnaissance report (paste inline -- agents do not share filesystem context automatically), the severity rubric and finding format from `references/shared-rubric.md`, and the agent-specific brief from its reference file. Each agent writes to its own output file under `audit-workspace/`.
   - Security -> `references/agent-security.md` -> `01-security-audit.md`
   - Performance -> `references/agent-performance.md` -> `02-performance-audit.md`
   - Accessibility -> `references/agent-accessibility.md` -> `03-accessibility-audit.md` (honor skip condition)
   - Dependency -> `references/agent-dependency.md` -> `04-dependency-audit.md` (honor skip condition)
   - Code Quality -> `references/agent-code-quality.md` -> `05-code-quality-audit.md`
6. For a targeted run (e.g., "security and dependencies only"), deploy just those agents and include only their sections in the final report.

### Phase 3: Report Compilation (sequential, ~5-10 min)

7. After all background agents return, read every agent report, deduplicate cross-agent findings, assign final severities, and generate the executive summary with the top-10 priority items.
8. Write the compiled `overnight-audit-report.md` to the repository root, then emit the brief completion message.

Follow `references/compilation-and-templates.md` for the deduplication map, final report template, and completion message.

## Scaling

| Codebase Size | Estimated Duration | Agent Strategy |
|---------------|-------------------|----------------|
| < 10K lines | 15-30 minutes | All agents, single pass each |
| 10K - 50K lines | 30-90 minutes | All agents, thorough pass |
| 50K - 200K lines | 1-4 hours | All agents, may need sub-agents for Security and Code Quality |
| 200K - 500K lines | 4-8 hours | All agents spawn 2-3 sub-agents each to parallelize file review |
| 500K+ lines | 8-14 hours | Full sub-agent deployment with file-batch assignments per sub-agent |

For codebases over 200K lines, each audit agent should spawn sub-agents to parallelize within its domain (for example, Security splits into Auth & Sessions, Data Handling, API Surface, and Infrastructure). Each brief instructs agents to self-organize sub-agent deployment based on the size discovered during their audit.

## Error Handling

- **Agent failure**: retry the agent once with the same brief. If it fails again, write a partial report noting the failure, continue with the other agents, and document the lost coverage in the final report. Never let one agent block the others.
- **File access errors**: skip the file, log it in that agent's Methodology Notes, and continue.
- **Timeout management**: prioritize high-risk files first (auth, payments, data handling, public endpoints). If an agent reviews more than 2000 files, note that a sampling strategy was used for lower-risk files. The Commander compiles partial results if the window is closing.
- **Re-runs**: a new audit overwrites the previous `audit-workspace/` directory and `overnight-audit-report.md`. Suggest the user commit or copy prior results first.

## Notes

- Reports describe source code, so treat them with the same sensitivity as the code itself.
- If `.gitignore` exists but does not list `audit-workspace/`, suggest adding it to prevent committing audit output.

© OneWave-AI · MIT · OneWave-AI/claude-skills

Similar skills

All skills for Code review and refactoring →
repo-publication-auditor davila7 repo-publication-auditor is a Claude Code subagent that audits a repository for exposed secrets in git history before public release. Runs before open-sourcing an internal project or flipping a private repository to public. CX Codex 53installs master-website-auditor Legion Code Inc. master-website-auditor — a skill that orchestrates a complete website audit through 20 specialized analysis modules: platform fingerprinting, content scanning, crawling, and parallel checks for SEO, security, accessibility, and performance. CX Codex +1 ★ 96repo stars ghost-repo-context ghostsecurity ghost-repo-context scans a codebase, finds projects, and writes an architecture map to repo.md. Use it before security analysis or when you need to orient yourself in a repository. CX Codex +1 4,200installs security-auditor Samarth Gupta security-auditor finds exploitable vulnerabilities in auth, untrusted-input handling, secrets, and dependencies. Triggered after changes to auth, API endpoints, or file operations. CX Codex ★ 276repo stars quality orwa-mahmoud This Claude Code skill groups project quality work into a single shift. It starts on requests about tests, accessibility, documentation, contracts, or security. CX Codex +1 ★ 571repo stars dependency-auditor alirezarezvani dependency-auditor is a skill that checks dependencies for vulnerabilities and license issues, then suggests fixes. Run it after dependency file changes or before deployment. CX Codex +1 ★ 788repo stars

From the same set185

The whole set →
customer-journey-mapper customer-journey-mapper generates an exhaustive customer journey document in customer-journey.md with seven lifecycle stages, Mermaid diagrams, and metrics. CX Codex +1 257installs contract-redliner contract-redliner reviews contracts clause by clause, rates risks, and drafts replacement language. Use it to prepare proposed edits and negotiation points. CX Codex +1 256installs deal-closer-playbook deal-closer-playbook analyzes in-flight B2B deals, runs web research on target accounts, and drafts an actionable deal-playbook.md document. CX Codex +1 256installs meeting-to-tasks meeting-to-tasks parses meeting transcripts into structured action items with assignees and deadlines, catches implicit commitments, and exports tasks for issue trackers. CX Codex +1 255installs cowork-deal-room The cowork-deal-room skill reviews deal-room files in five stages and produces risk findings, term comparisons, and negotiation guidance. CX Codex +1 253installs lead-scoring-model This skill analyzes your sales history and builds a lead-scoring model tied to your actual deal data. It generates documentation with scoring dimensions, thresholds, and a CRM implementation guide. CX Codex +1 253installs incident-responder The incident-responder skill investigates production incidents by checking logs and recent changes, classifying severity, and preparing recovery guidance and a report. Use it during an outage and for follow-up. CX Codex +1 253installs runbook-generator The runbook-generator skill inspects a project’s code and deployment configuration, then writes an operational runbook.md. Use it to document releases, incident response, or disaster recovery. CX Codex +1 252installs

Project sponsors

Companies that keep this analytics open