octocode-cli
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses octocode-cli, jq, and printf to interact with GitHub and process data.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface.
- Ingestion points: Data is retrieved from GitHub via octocode-cli search-code and get-file.
- Boundary markers: No delimiters are used to isolate untrusted content.
- Capability inventory: The agent can execute shell commands based on retrieved data.
- Sanitization: No validation of external content is performed.
Audit Metadata