Index of every page on this site: llms.txt
All skills
openai avatar

/security-scan

@ec0d6c1
by openaiopenai/codex-security14k stars
1,108
Run it yourself

Find vulnerabilities in the current source of a whole repository or selected paths. This is the default single-pass audit. Use the diff skill for PR, commit, branch, or working-tree changes and the deep skill for an explicitly requested deep scan.

Use this Skill: https://skilld.dev/gh/openai/codex-security/security-scan

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesscan-results.md

≈718 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Save results to the registered scan

When the SDK supplies only a registered environment, use CODEX_SECURITY_REPOSITORY as the target. If CODEX_SECURITY_TARGET_PATHS_FILE is set, use the supplied source inventory or resolve it before review. Emit the SDK's standalone CODEX_SECURITY_SCAN_PROGRESS {"phase":"discovery","filesCompleted":3,"filesTotal":8} at discovery start, meaningful review batches, and later phase transitions using the actual phase, distinct completed paths and scoped total or host estimate.

For a scan with bound progress and context tools, persist user-requested context edits with the current handoff token. Use each real forward phase transition’s returned scan.userContext for that phase and its workers. Without a bound context tool, retain the original context.

At the first usable threat model, call record_codex_security_scan_draft with the scan ID, handoff token when required, complete: false, the model, and empty findings if none are validated yet. The workbench writes the canonical artifacts and derives target metadata and finding identities. If the SDK supplied the scan and no draft tool is bound, follow the SDK file handoff through completion instead. Other hosts require the draft tool.

Checkpoint returned candidates before merging or validating them. Keep their original evidence under reviewState.deferred until resolved. Reuse saved surface IDs. Preserve validation decisions, counterevidence, and unfinished questions as directed by the core audit. Use artifact storage for supplemental evidence. A caller-supplied custom validation handoff determines which candidates are provisional.

When assigned work finishes, submit one accepted complete: true draft with scope, threat model, findings, and reviewState. Close finished generic tasks in reviewState.resolvedDeferred with their saved ID and a reason. Keep unresolved work. For rejected drafts and interrupted writes, follow Errors And Recovery.

Verify scan-manifest.json and findings.json exist. For an SDK-owned scan, return now. The SDK seals the results and generates the reports. For another managed scan, follow any desktop handoff if on desktop, then call complete_codex_security_scan once with the same ID and token. On success, link the generated report.md and canonical artifacts. Retrieve full findings only if requested. Report only complete measured usage. Leave pending usage to the host and identify other incomplete counts as unavailable.

If completion fails, surface the error and preserve the durable scan. Cancel only at the user's request. Fail the scan only for a confirmed unrecoverable blocker after recovery is exhausted. Leave unfinished work resumable.

Source: SKILL.md on GitHub

No rule matched.

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No alerts8d3 checks · Risk SAFE
  • Gen Agent Trust Hub8d

    This skill facilitates automated security audits by utilizing local scripts to analyze repository source code. It includes built-in safeguards, such as explicit instructions to treat input files as data and to avoid following external links, which help maintain a secure analysis environment. The use of platform-specific commands is consistent with the skill's role as a cross-platform development tool.

  • Socket9d

    No alerts

  • Snyk8d

    Risk: LOW · No issues

Signed by skilld at ec0d6c1. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated last week

README badge

README badge for openai/codex-security/security-scan