github-workflow

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill explicitly retrieves and executes commands against external task sources like ClickUp/Jira (see references/feature-task-retrieval.md and the ~/.bonfire/source.json command examples) so it ingests untrusted, user-generated task and comment content which the agent reads and uses to generate TODOs, delegate work, and drive commits/PRs.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill runs "npx skills find " at runtime, which fetches and executes the remote "skills" npm package from the npm registry (e.g. https://registry.npmjs.org/skills), allowing remote code to run and influence discovery results/agent behavior.