deep-security-scan — Codex skill: full repository security audit
This skill runs multiple independent security scans across your entire repository or a scoped path, aggregates findings, and generates a complete report. Use it when standard scanning needs more depth and confidence.
npx skills add openai/plugins --skill deep-security-scan -a codexWhat deep-security-scan does
deep-security-scan uses the Codex Security tool to perform repeated independent audits. Each worker reads the reference scan config and completes a full audit; the coordinator collects findings and produces scan-manifest.json, findings.json, and coverage.json. The skill does not work on pull requests, commits, or diffs—only full repository or directory scans.
Before starting, the skill checks for concurrent deep scans, since running two at once slows both and increases CPU and memory use. If you supply a scan context with URLs or directory exclusions, it passes them to all workers as immutable data and reads each source exactly once.
Install via npx skills add openai/plugins --skill deep-security-scan (Codex) or platform-specific commands for Claude Code, GitHub Copilot, Cursor, Gemini CLI, and Windsurf. The skill config appears in your project after install: .agents/skills/, ~/.claude/skills/, .cursor/skills/, .github/skills/, or .devin/skills/.
What it helps with
- run multiple independent security scans to catch more vulnerabilities
- aggregate and validate findings from parallel audit workers
- generate a full coverage report with canonical artifacts
Example prompts for the agent
- Run a deep security scan of the entire repository
- Scan only the `src/` directory and report all findings
- Audit the full project, exclude `node_modules` and `build`, and merge results from multiple passes
How to install deep-security-scan
1Skills CLI (Vercel)
npx skills add openai/plugins --skill deep-security-scan -a codex2Manually: put the files in place
.agents/skills/deep-security-scan/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill deep-security-scan -a claude-code2Manually: put the files in place
~/.claude/skills/deep-security-scan/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/deep-security-scan/
Manually: put the files in place
.cursor/skills/deep-security-scan/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins deep-security-scan2Manually: put the files in place
.github/skills/deep-security-scan/SKILL.mdPlatform CLI
gemini skills install https://github.com/openai/plugins --path plugins/codex-security/skills/deep-security-scanManually: put the files in place
.devin/skills/deep-security-scan/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set201
The whole set →Similar
Questions about deep-security-scan
How do I install deep-security-scan for Codex?
Run npx skills add openai/plugins --skill deep-security-scan, then npx skills add openai/plugins --skill deep-security-scan -a codex. The skill config is at .agents/skills/deep-security-scan/SKILL.md.
What does deep-security-scan do differently from a standard scan?
A standard Security scan makes one pass. deep-security-scan launches multiple independent workers that each complete a full audit, then aggregates and validates the combined findings. This catches more issues and increases confidence in the results.
Can I scan a specific folder instead of the entire repo?
Yes. Provide a scoped path and the skill scans only that directory and its contents, applying the same multi-pass validation approach.
