sky
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches knowledge base content and management scripts from the author's official GitHub repository (github.com/arcniko/sky-kb.git).
- [REMOTE_CODE_EXECUTION]: Executes a Python synchronization script (sync.py) downloaded from the author's repository.
- [COMMAND_EXECUTION]: Utilizes bash commands including git, mkdir, and rm -rf to manage the local knowledge base directory and files.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes data from external repositories. 1. Ingestion points: Reads DIRECTORY.md, .kb_config.json, and content files from cloned repositories. 2. Boundary markers: No explicit instructions are provided to the agent to ignore instructions embedded within the knowledge base content. 3. Capability inventory: The skill can execute bash commands (git, python3, rm -rf), write files, and modify its own configuration. 4. Sanitization: No input validation or content filtering is performed on the ingested data.
Audit Metadata