We take security seriously. This document covers how to report a vulnerability in ks-cookbook, the knowledgestack-mcp server, or any code distributed from this repository.
Only the main branch receives security updates. Pinned or tagged releases are best-effort — please upgrade to the latest before filing a report.
| Version | Supported |
|---|---|
main |
✅ |
| tagged releases | |
| forks / patches | ❌ |
Do not open a public GitHub issue for security vulnerabilities.
Please report privately via one of:
- GitHub Private Vulnerability Reporting (preferred) — on this repo, click Security → Report a vulnerability. This creates a private advisory only visible to maintainers. Link: report here.
- Email —
security@knowledgestack.ai(if you don't use GitHub). Please include "ks-cookbook" in the subject.
Include, where possible:
- Affected component (flagship name,
knowledgestack-mcp, a recipe, etc.) - Reproduction steps (minimum viable command or diff)
- Impact (data exposure, code execution, privilege escalation, denial of service)
- Affected commit / version
- Any proof-of-concept you're willing to share
Examples of things we do want to hear about:
- Secret/credential exposure in example code or logs
- Prompt-injection or tool-invocation vulnerabilities in the MCP server itself (not in downstream agent code — that's the user's responsibility)
- Dependency vulnerabilities with a clear exploit path in our code
- Supply-chain concerns in our published artifacts (
knowledgestack-mcpon PyPI, generated SDKs) - CI/CD or GitHub Actions misconfigurations that could allow unauthorized writes
Things that are not security issues here:
- Bugs in your own agent logic that leak data (that's an agent-level concern; file a normal issue if you need help)
- Vulnerabilities in the hosted Knowledge Stack API — file those privately at security@knowledgestack.ai; they don't belong in this repo
- Prompt-injection in flagships you've forked and modified (we can't triage unbounded variants)
- Acknowledgement: within 2 business days of your report
- Triage & severity assessment: within 5 business days
- Fix timeline: depends on severity — critical issues target 7 days; lower-severity issues within the next release cycle
- We will coordinate on disclosure timing — default is 90 days from report, with mutual agreement on earlier or later publication if warranted
Security research conducted in good faith against this repository and the referenced MCP server is welcomed. We won't pursue legal action against researchers who:
- Avoid privacy violations, destruction of data, and disruption of service
- Only interact with accounts they own or have explicit permission to test
- Give us a reasonable window to fix before publicly disclosing
- Do not monetize the vulnerability before disclosure
Thanks for helping keep the Knowledge Stack ecosystem safe.