fix-finding — Codex skill: what it does, how to install
Resolves confirmed security vulnerabilities through targeted, validated patches. Useful when you need to close an exploit path without breaking existing APIs.
npx skills add openai/plugins --skill fix-finding -a codexWhat fix-finding does
This skill turns a specific vulnerability report into a minimal code fix. It begins by tracing the data path from input to sink. When delegation is available, an isolated read-only subagent inspects the security boundary. If the repository already protects the path, the skill reports no_change and avoids speculative edits.
After the patch is written, a second review subagent challenges the implementation. It looks for alternative bypasses and checks whether standard inputs now fail. The skill also handles Generate and Apply remediation stages within the Codex Security workbench.
Skip this skill for standard bug fixes or general code cleanup. Install it only when you handle explicit vulnerability remediation.
What it helps with
- Trace tainted inputs from source to sink across repository boundaries
- Generate minimal security fixes without altering unrelated functions
- Verify patches against alternative bypass routes and edge cases
- Prepare isolated unified diffs for Codex Security remediation flows
Example prompts for the agent
- Fix this path traversal vulnerability and verify that existing file tests still pass.
- Remediate the SSRF finding in our webhook handler with a minimal patch.
- Generate a validated diff for the authorization bypass occurrence in the auth controller.
How to install fix-finding
1Skills CLI (Vercel)
npx skills add openai/plugins --skill fix-finding -a codex2Manually: put the files in place
.agents/skills/fix-finding/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add openai/plugins --skill fix-finding -a claude-code2Manually: put the files in place
~/.claude/skills/fix-finding/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/fix-finding/
Manually: put the files in place
.cursor/skills/fix-finding/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install openai/plugins fix-finding2Manually: put the files in place
.github/skills/fix-finding/SKILL.mdPlatform CLI
gemini skills install https://github.com/openai/plugins --path plugins/codex-security/skills/fix-findingManually: put the files in place
.devin/skills/fix-finding/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 fix-finding
How to install fix-finding in Codex?
Run npx skills add openai/plugins --skill fix-finding -a codex in your terminal. This adds the skill definition to .agents/skills/fix-finding/SKILL.md.
What does fix-finding do?
It investigates reported security issues, implements a minimal fix at the enforcement boundary, and validates that normal behavior stays intact.
When should you avoid using this skill?
Do not use it for standard code review, routine bug fixing, or broad repository scans. It activates only for confirmed or plausible security findings.