xy-coach
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill uses a Python script (
scripts/atoms-search.py) to connect to the vendor-controlled API athttps://api.xyskill.xyzto retrieve knowledge base "atoms". - [COMMAND_EXECUTION]: The skill instructions in
SKILL.mddirect the agent to execute a shell script (scripts/xy-init.sh) for initialization and a Python script (scripts/atoms-search.py) for searching the knowledge base. - [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied information to build a business profile and uses user keywords in command-line search queries, creating a potential injection surface.
- Ingestion points: User input is ingested during the profile creation process and subsequent coaching interactions (SKILL.md).
- Boundary markers: The skill lacks explicit instructions for the agent to sanitize or delimit user input before passing it to internal scripts.
- Capability inventory: The skill has the capability to execute bash and Python scripts with arguments derived from user interaction.
- Sanitization: The
atoms-search.pyscript performs basic string splitting but does not sanitize for shell-specific special characters, relying on the host agent's execution environment.
Audit Metadata