cicd-security
Skill1 starsChanged 3 months agoMITView on GitHub
CI/CD pipeline security hunting — GitHub Actions workflow injection, secret exfiltration, self-hosted runner poisoning, dependency confusion, OIDC token theft, and supply chain attacks. Covers sisakulint scanning, manual workflow analysis, and chaining CI/CD bugs into critical findings. Use when a target has public repos, GitHub Actions, CircleCI, Jenkins, or GitLab CI.
- Reads credentials
6 other projects use a version of this, 3 of them edited.
Adopted by
Most recently taken up by
- thesammykins/skillstheir own version, first seen Nov 1, 2026
- briiirussell/cybersecurity-skillstheir own version, first seen Nov 1, 2026
- shuvonsec/claude-bug-bountyedited, first seen Oct 8, 2026
- conjure-3301/skillstheir own version, first seen Aug 17, 2026
- awarexone/Agentic-Bug-Hunteredited, first seen Jun 9, 2026
- Awarexone/Agentic-Bug-Hunteredited, first seen Jun 9, 2026
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 3
- shuvonsec/claude-bug-bounty★ 5.9k21% same
- awarexone/Agentic-Bug-Hunter★ 5.9k21% same
- Awarexone/Agentic-Bug-Hunter★ 5.9k21% same
Wrote their own version · 3
- briiirussell/cybersecurity-skills★ 489same name
- conjure-3301/skills★ 0same name
- thesammykins/skills★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository50 other files it gives its agents.
- AGENTS.mdAGENTS.md
- .claude/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- agent-safetySkill
- ai-llmSkill
- apiSkill
- asset-graphSkill
- authSkill
- auto-researchSkill
- binary-analysisSkill
- business-logicSkill
- campaignSkill
- cloudSkill
- cors-csrfSkill
- coverageSkill
- cross-accountSkill
- domain-modelSkill
- enrichmentSkill
- evaluation-harnessSkill
- file-uploadSkill
- graphql-auditSkill
- http-protocolSkill
- identity-domainSkill
- impact-verifierSkill
- meme-coin-auditSkill
- mobileSkill
- modern-browserSkill
- nuclei-scannerSkill
- oob-infraSkill
- osintSkill
- personaSkill
- plannerSkill
- privescSkill
- program-memorySkill
- race-conditionSkill
- rceSkill
- reconSkill
- reportingSkill
- sastSkill
- scope-managerSkill
- skill-scientistSkill
- sqliSkill
- ssrfSkill
- standard-catalogSkill
- technique-kbSkill
- traffic-corpusSkill
- triage-validationSkill
- vuln-intelSkill
- web3-auditSkill
- xssSkill
The file
Files
CI/CD SECURITY — Pipeline Attack Surface
Overview
CI/CD pipelines are high-value targets — a single workflow injection can give you code execution on the build server, read ALL org secrets, and push backdoored releases to production.
This skill hunts CI/CD vulnerabilities in public repositories: script injection
through GitHub Actions contexts, pull_request_target misuse, secret
exfiltration paths, self-hosted runner poisoning, OIDC token abuse, dependency
confusion, and unpinned actions. Workflow scan-workflows runs the automated
scan; the runbooks cover manual analysis of each bug class.
Quick Reference
Quick kill checklist
[ ] Run cicd_scanner.sh <owner/repo> — catch low-hanging workflow lint issues
[ ] Check for script injection: ${{ github.event.*.body/title/name }}
[ ] Find secrets referenced in env: — test if they leak in logs
[ ] Check pull_request_target with checkout of untrusted code
[ ] Look for self-hosted runners on public repos
[ ] Search for OIDC token requests without audience restriction
[ ] Check for unpinned actions (uses: owner/action@main)
[ ] Look for workflow_dispatch with no input validation
[ ] Find artifact downloads without integrity checks
[ ] Search for GITHUB_TOKEN with write permission used insecurelyInjectable contexts (always check these)
github.event.pull_request.title
github.event.pull_request.body
github.event.pull_request.head.ref ← branch names
github.event.issue.title
github.event.issue.body
github.event.comment.body
github.event.review.body
github.event.review_comment.body
github.event.discussion.title
github.event.discussion.body
github.head_ref ← alias for branch name
github.event.inputs.* ← workflow_dispatch inputsBug class table
| Bug | Trigger | Severity | CVSS Range |
| Workflow injection via PR title | ${{ github.event.pull_request.title }} in run: | Critical | 9.0–10.0 |
pull_request_target + checkout | Accepts PRs from forks | Critical | 9.0–10.0 |
| Self-hosted runner on public repo | runs-on: self-hosted + public repo | High | 7.5–9.0 |
| OIDC trust too broad | Any-branch/any-repo claim | High | 7.5–8.5 |
| Secret in log | echo ${{ secrets.X }} | Medium | 5.5–7.0 |
| Unpinned action | @main / @v1 tag | Low–Medium | 3.0–5.5 |
| Artifact poisoning | Unsigned artifact download + exec | Medium | 5.5–7.0 |
GITHUB_TOKEN write abuse | Push to protected branch | Medium | 5.5–7.0 |
| Dependency confusion | Internal pkg not on public registry | High | 7.5–9.0 |
workflow_dispatch injection | Unvalidated inputs in run: | Medium–High | 6.0–8.0 |
Workflow Selection
| Situation | Workflow |
| Scan a public repo/org's workflows | scan-workflows |
| Local checkout available — static injection scan | scripts/workflow_inject_scan.py |
| Look for hardcoded secrets in CI configs | scripts/cicd_secret_scan.py |
| Deep dive on a found pattern | matching runbook (injection / pull-request-target / secrets / runners) |
Available Workflows
scan-workflows
Scans GitHub Actions workflows for a repo or org via sisakulint (remote scan), then runs the local static injection and secret scans on any local checkout.
bash .claude/skills/cicd-security/scripts/cicd_scanner.sh "${REPO:?Set REPO (owner/repo)}" \
--output-dir $OUTDIR/cicdOutputs: $OUTDIR/cicd/scan_results.txt, $OUTDIR/cicd/summary.txt. Safety tier: passive (reads public workflow files only).
Evidence Required
| Artifact | File |
| Workflow file with the vulnerable pattern | $OUTDIR/cicd/scan_results.txt |
| Proof of injectable context (grep line) | $OUTDIR/cicd/evidence/ |
| PoC payload used (PR title / issue title) | evidence/<finding>/poc.sh |
| Public repo + workflow path + line number | in finding report |
References
- Workflow injection deep dive:
runbooks/workflow-injection.md pull_request_targetmisuse:runbooks/pull-request-target.md- Secret exfil patterns:
runbooks/secret-exfiltration.md - Self-hosted runner poisoning:
runbooks/runner-poisoning.md - Scope notes: most programs scope public repos only; opening a real PR to trigger a workflow requires explicit program permission
- Redaction policy: scan output stays local-only under
$OUTDIR; redact any secret values surfaced by scans before committing evidence.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

