Traces a security finding from source to sink and calibrates its severity.
openai/codex-security
Finds, validates, fixes and tracks security vulnerabilities in code using OpenAI's Codex Security tools.
- 20 skills
- 11.4k stars
- Reached #2 on GitHub's daily trending
The 5 most installed
Ordered by installs on skills.sh- attack-path-analysis813 installs
- threat-model203 installs
Creates, updates or saves a repository threat model.
- security-scan157 installs
Finds vulnerabilities in a whole repository or selected paths in a single default audit pass.
- deep-security-scan127 installs
Runs repeated security scans over a repository or path for an exhaustive multi-pass audit.
- propose-security-hardening117 installs
Builds evidence-backed proposals for structural security improvements beyond individual patches.
The other 15 skills
- security-diff-scan116 installs
Reviews a pull request, commit, branch diff or working-tree patch for security vulnerabilities.
- define-security-policy113 installs
Writes or updates SECURITY.md guidance on what should be reviewed and which security properties must hold.
- fix-finding112 installs
Fixes and verifies a validated or plausible security vulnerability when you explicitly ask.
- track-findings103 installs
Tracks validated security findings in Linear, Jira, GitHub issues or draft security advisories.
- triage-finding103 installs
Triages supplied or imported security findings against a repository's policy and code evidence.
- validation102 installs
Determines whether one or more candidate security findings are actually valid.
- finding-discovery101 installs
Searches a repository or code change for candidate security findings.
- verify-fix89 installs
Verifies that a security fix actually remediates a reported vulnerability when you explicitly ask.
- assess-patch-risk78 installs
Reviews a patch's impact, regression risk and auto-merge eligibility without changing the patch.
- cloud-findings40 installs
Reads existing Codex Security Cloud findings, reports and scan status from your connected account.
- publish-findings33 installs
Imports findings from scanners, CI reports and vendor tools, and publishes selected ones to Codex Security Cloud.
- check-skills30 installs
Checks an agent skill's files, scripts and permissions for unsafe instructions before you install it.
- resolve-dependency-alert30 installs
Investigates a vulnerable dependency alert, traces what pulls it in and checks whether it can be upgraded.
- threat-model-visualize30 installs
Turns an existing threat model into an interactive architecture and threat-path explorer.
Turns vulnerability notes, reports, PoCs or findings into self-contained vulnerability reports.