ark-vulnerability-fixer
Skill439 starsChanged 7 months agoApache-2.0View on GitHub
CVE research and security patch workflow for Ark. Provides CVE API integration, mitigation strategies, and security-focused PR templates. Works with research, analysis, and setup skills for comprehensive vulnerability fixing.
- Installs packages
- Commits and pushes
In this repository29 other files it gives its agents.
- ark/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- tests/CLAUDE.mdCLAUDE.md
- a2a-protocolSkill
- Ark AnalysisSkill
- ark-architectureSkill
- ark-controller-developmentSkill
- ark-dependabot-managementSkill
- ark-sdk-developmentSkill
- ark-setup-clusterSkill
- ark-chainsaw-testingSkill
- Ark Dashboard and UI TestingSkill
- ark-documentationSkill
- issue-creationSkill
- ark-issuesSkill
- multi-branch-cluster-setupSkill
- openspec-apply-changeSkill
- openspec-archive-changeSkill
- openspec-bulk-archive-changeSkill
- openspec-continue-changeSkill
- openspec-exploreSkill
- openspec-ff-changeSkill
- openspec-new-changeSkill
- openspec-onboardSkill
- openspec-sync-specsSkill
- openspec-verify-changeSkill
- ark-pentest-issue-resolverSkill
- ark-researchSkill
- slack-announcementSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
17 rules. About 0 are common, 15 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈15
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 12
Use this skill when:
- Distinctive: line 57
Tip: Use the research skill for web searches to find vendor advisories and GitHub security alerts.
- Distinctive: line 85
Consider Ark's specific context:
- Distinctive: line 91
Tip: Use the analysis skill to understand Ark's architecture and service boundaries.
- Distinctive: line 97
CRITICAL: Always present mitigation options and wait for user approval before making changes.
- Distinctive: line 99
Use this template to present findings:
- Distinctive: line 154
After user approves the mitigation, clone Ark for making changes:
- Distinctive: line 237
Only updating documentation or CLI
- Distinctive: line 245
Ensure you're in the cloned repository:
- Distinctive: line 290
Create the PR with detailed security information:
- Distinctive: line 372
Test integration (this skill + setup skill, if needed)
- Distinctive: line 373
Create PR (this skill)
- Very specific: line 399
npm overrides: Use "overrides" in package.json to force a patched version.
- Distinctive: line 400
Version conflict resolution: When a transitive dep has incompatible version requirements, consider upgrading the parent package instead.
- Distinctive: line 404
Update: Edit requirements files
- Very specific: line 408
Check: Dockerfile FROM statements
- Distinctive: line 409
Update: Change base image version
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Ark Vulnerability Fixer
Provides CVE-specific research tools and security patch workflows for fixing vulnerabilities in Ark.
When to use this skill
Use this skill when:
- User mentions a specific CVE number (e.g., "Fix CVE-2025-55183 in Ark")
- User reports a security vulnerability that needs patching
- You need CVE database information
- You need security-focused PR templates
Note: This skill is typically used by the ark-security-patcher agent as part of a complete workflow:
- CVE research (this skill + research skill)
- Codebase analysis (this skill + analysis skill)
- Mitigation planning (this skill)
- Repository cloning and fix implementation
- Testing (optionally with setup skill)
- PR creation (this skill)
This skill complements the research, analysis, and setup skills for a complete end-to-end vulnerability fixing workflow.
CVE Research
CVE API Integration
Fetch official CVE data from the CIRCL CVE database:
# Fetch CVE details
curl -s "https://cve.circl.lu/api/cve/CVE-2025-55183" | python3 -m json.toolThe API provides:
- Official CVE description
- CVSS scores and severity ratings
- References to security advisories
- Affected products and version ranges
- CWE categorization
- Available patches and fixes
CVE Research Checklist
For each CVE, gather:
- Official CVE description and CVSS score
- Vendor security advisory
- GitHub security advisory (if applicable)
- Patch or fix documentation
- Affected version range
- Recommended version or workaround
Tip: Use the research skill for web searches to find vendor advisories and GitHub security alerts.
Dependency Analysis
Identifying Vulnerable Dependencies
Once you have CVE details, search Ark's dependencies:
cd /tmp/ark-analysis # Use analysis skill to clone first
# Go dependencies
grep "package-name" go.mod go.sum
go list -m all | grep "package-name"
# Node.js dependencies
find . -name "package.json" -exec grep -l "package-name" {} \;
npm list package-name # If in a node project
# Python dependencies
find . -name "requirements.txt" -o -name "pyproject.toml" | xargs grep "package-name"
# Docker base images
find . -name "Dockerfile" | xargs grep "FROM"Assessing Impact
Consider Ark's specific context:
- Deployment model: Kubernetes operator in cluster
- Network exposure: Services typically internal to cluster
- Trust boundary: Often in trusted environments
- Attack vectors: What's realistic given Ark's architecture?
Tip: Use the analysis skill to understand Ark's architecture and service boundaries.
Mitigation Strategy
Presenting Options to User
CRITICAL: Always present mitigation options and wait for user approval before making changes.
Use this template to present findings:
## Security Vulnerability Analysis
### Vulnerability Details
- **CVE**: CVE-YYYY-NNNNN (or "Generic: [description]")
- **Severity**: [Critical/High/Medium/Low] (CVSS: [score])
- **Component**: [Library/package/framework]
- **Description**: [Clear explanation]
### Impact on Ark
- **Affected Services**: [List services/components]
- **Current Version**: [Version in use]
- **Vulnerable Versions**: [Range]
- **Attack Vector**: [How exploitable]
- **Risk Assessment**: [Realistic risk for Ark deployments]
### Mitigation Options
#### Option 1: [Recommended approach] (RECOMMENDED)
- **Action**: Update [component] from v[X] to v[Y]
- **Changes Required**: [Files to modify]
- **Testing Strategy**: [How to verify]
- **Impact**: [Breaking changes, if any]
- **Pros**: [Benefits]
- **Cons**: [Downsides]
#### Option 2: [Alternative approach]
- **Action**: [Alternative fix]
- **Changes Required**: [What changes]
- **Testing Strategy**: [How to verify]
- **Impact**: [Breaking changes, if any]
- **Pros**: [Benefits]
- **Cons**: [Downsides]
### Recommendation
Based on [evidence sources], I recommend **Option 1** because:
1. [Primary reason]
2. [Secondary reason]
### Next Steps
Would you like to proceed with this mitigation?
### Sources
- [CVE Database](https://cve.circl.lu/cve/CVE-YYYY-NNNNN)
- [Vendor Advisory](URL)STOP AND WAIT for user approval before implementing.
Repository Setup for Fixes
Cloning for Development
After user approves the mitigation, clone Ark for making changes:
# Clone the repository
git clone git@github.com:mckinsey/agents-at-scale-ark.git
cd agents-at-scale-ark
# Create a security fix branch
git checkout -b security/fix-cve-YYYY-NNNNN
# Verify branch
git branch --show-currentFor forks:
git clone git@github.com:<username>/agents-at-scale-ark.git
cd agents-at-scale-ark
git remote add upstream git@github.com:mckinsey/agents-at-scale-ark.git
git fetch upstream
git checkout -b security/fix-cve-YYYY-NNNNN upstream/mainImplementation
Applying the Fix
Once user approves and repository is cloned, apply changes:
cd agents-at-scale-ark
# For Go dependencies
go get package@v1.2.3
go mod tidy
# For Node.js dependencies
npm install package@1.2.3
npm audit fix
# For Python dependencies
# Edit requirements.txt or pyproject.toml
pip install -r requirements.txt
# For Docker base images
# Edit Dockerfile FROM statementsVerification
Basic Testing
cd agents-at-scale-ark
# Run tests
make test
# Build to check for breaking changes
make build
# Search for remaining vulnerable patterns
grep -r "vulnerable-pattern" .Integration Testing with Setup Skill (Optional)
For changes that affect Ark runtime behavior, use the setup skill to test in a live cluster:
When to use setup skill for testing:
- Go operator changes (controllers, webhooks, CRDs)
- Service updates (ark-api, executor services)
- Changes that affect Kubernetes interactions
- Breaking changes that need verification
Setup skill workflow:
- Creates a Kind cluster
- Builds ark-cli from your security fix branch
- Installs Ark with your changes
- Verifies all pods are running
- Allows you to test the fix in action
Skip integration testing if:
- Only updating documentation or CLI
- Changes are in isolated utility functions
- Dependencies don't affect runtime behavior
Security-Focused PR Templates
Commit Message Template
Ensure you're in the cloned repository:
cd agents-at-scale-ark
git add .
git commit -m "$(cat <<'EOF'
fix: CVE-YYYY-NNNNN in [component]
## Vulnerability Details
- CVE: CVE-YYYY-NNNNN
- Severity: [Critical/High/Medium/Low]
- CVSS Score: [X.X]
- Component: [package/library]
## Impact on Ark
[How this affects Ark services and realistic risk level]
## Changes
- Updated [component] from v[X] to v[Y]
- [Any code changes]
## Testing
- [Tests run and results]
## References
- CVE: https://cve.circl.lu/cve/CVE-YYYY-NNNNN
- Advisory: [URL]
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
EOF
)"Push to Remote
# Push the security fix branch
git push origin security/fix-cve-YYYY-NNNNNPull Request Template
Create the PR with detailed security information:
gh pr create --title "fix: CVE-YYYY-NNNNN in [component]" --body "$(cat <<'EOF'
## Summary
Addresses security vulnerability CVE-YYYY-NNNNN in [component].
## Vulnerability Details
| Field | Value |
|-------|-------|
| **CVE** | CVE-YYYY-NNNNN |
| **Severity** | [Critical/High/Medium/Low] |
| **CVSS Score** | [X.X] |
| **Component** | [package] |
| **Current Version** | [old] |
| **Patched Version** | [new] |
### Description
[What the vulnerability is and how it could be exploited]
## Impact on Ark
### Affected Components
- [Service 1]: [Impact]
- [Service 2]: [Impact]
### Risk Assessment
**Risk Level**: [Level]
[Realistic assessment of actual risk to Ark deployments]
## Changes Made
- Updated `[component]` from `v[X]` to `v[Y]`
- [Other changes]
## Testing
- ✅ Unit tests pass
- ✅ Integration tests pass
- ✅ Manual verification completed
## References
- **CVE**: https://cve.circl.lu/cve/CVE-YYYY-NNNNN
- **Advisory**: [URL]
- **Patch Notes**: [URL]
🤖 Generated with [Claude Code](https://claude.com/claude-code)
EOF
)"Important Notes
CVE API Usage
The CIRCL CVE API:
- Endpoint:
https://cve.circl.lu/api/cve/{CVE-ID} - Returns JSON with CVSS scores, references, affected versions
- No authentication required
- Fallback: Use web search if API is unreachable
Ark Security Context
When assessing risk:
- Architecture: Kubernetes operator managing AI workloads
- Components: Go operator, Python services, Node.js CLI
- Deployment: Typically cluster-internal, trusted environments
- Focus areas: CRD controllers, API services, executor services
Skill Composition
This skill provides CVE-specific tools. It works best when combined with:
- research skill - For web searches, vendor advisories, evidence gathering
- analysis skill - For cloning Ark repo (read-only) and examining codebase structure
- setup skill - For integration testing in a live Ark cluster
- architecture skill - For understanding service boundaries and impact
Complete workflow example:
- Research CVE (this skill + research skill)
- Analyze impact (this skill + analysis skill)
- Clone for development (this skill)
- Implement fix (this skill)
- Test integration (this skill + setup skill, if needed)
- Create PR (this skill)
User Approval is Mandatory
Never implement changes without explicit user approval. This ensures:
- User understands security implications
- Approach aligns with security policies
- Testing strategy is appropriate
- Breaking changes are acknowledged
Common Vulnerability Types
Go Dependencies
- Check:
go.mod,go.sum - Update:
go get package@version && go mod tidy - Scan:
go list -m all
Node.js Dependencies
- Check:
package.json,package-lock.json - Update:
npm install package@version - Scan:
npm audit - Transitive deps: A vulnerable package may appear at multiple versions across nested
node_modules/. Check ALL lockfiles:find . -name "package-lock.json" -exec grep -l "<package>" {} \; jq -r '.packages | to_entries[] | select(.key | contains("<package>")) | "\(.key)\t\(.value.version)"' <lockfile> - npm overrides: Use
"overrides"inpackage.jsonto force a patched version. But overrides are global — they affect ALL consumers of that package. If two consumers need incompatible major versions (e.g., express 4.x needspath-to-regexp ~0.1.xwhile router needs^8.x), an override will break one of them. - Version conflict resolution: When a transitive dep has incompatible version requirements, consider upgrading the parent package instead. Example: upgrading express 4.x to 5.x unified
path-to-regexpto a single 8.x version, resolving the conflict without overrides.
Python Dependencies
- Check:
requirements.txt,pyproject.toml - Update: Edit requirements files
- Scan:
pip-audit(if available)
Docker Base Images
- Check:
DockerfileFROM statements - Update: Change base image version
- Scan:
docker scanor vulnerability databases
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.

