library-add
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes several local scripts and tools to process and categorize content, including
./scripts/library-add.py,./scripts/skills-db.sh, andtools/ai-lib/ai-lib. These scripts have the capability to write to and modify the local repository structure. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection where untrusted data could influence the agent or the repository state.
- Ingestion points: Processes user-provided "pasted text" or files specified via the
--fileargument inSKILL.md. - Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions embedded within the content being added to the library.
- Capability inventory: The skill uses
./scripts/library-add.pyand./scripts/skills-db.shto write new files to the filesystem (e.g., inskills/orprompts/) and rebuild local catalogs. - Sanitization: No validation, sanitization, or escaping of the user-provided content is mentioned before it is processed and written to disk.
Audit Metadata