pylon
Warn
Audited by Snyk on Feb 25, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). The skill's required workflow explicitly pulls human-authored review content and threaded comments from the Pylon web UI (see SKILL.md Quick Workflow and references/tools-plan.md / references/tools-code-review.md: pull_plan and pull_code_feedback), which the agent reads and uses to decide next actions, so untrusted third-party (user-generated) input can indirectly inject instructions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 1.00). The skill requires calling a configured MCP endpoint (e.g. https://your-pylon-instance.com/api/mcp or http://localhost:3000/api/mcp) at runtime to pull human-reviewed plans/decisions which directly control the agent's subsequent instructions, so this external URL is a runtime dependency that controls prompts.
Audit Metadata