octocode-cli

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

• Third-party content exposure detected (high risk: 1.00). This skill's workflow (SKILL.md) explicitly instructs the agent to run octocode-cli commands like search-code and get-file against public GitHub repositories (owner/repo paths) and to read those search results and file contents to drive follow-up searches and decisions, which exposes it to untrusted, user-generated third-party content from public repos.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs octocode-cli to fetch repository files at runtime (e.g., https://github.com/bgauryy/octocode-mcp), and those fetched file contents are read and used to form answers—i.e., external repository content is injected into the agent context and thus can directly control prompts.