xy-peer-pick
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes a bundled script (
scripts/atoms-search.py) to query an external knowledge base athttps://api.xyskill.xyz/v1/atoms/searchduring runtime. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data retrieved from a remote API and interpolates it into the agent's context without explicit sanitization or boundary markers, creating a surface for potential injection attacks if the remote source is compromised.
- Ingestion points:
SKILL.mdinstructions utilize the output from thescripts/atoms-search.pytool, which fetches data fromapi.xyskill.xyz. - Boundary markers: Absent. The instructions do not provide delimiters or "ignore embedded instructions" warnings for the tool's output.
- Capability inventory: The skill has the ability to execute local scripts and perform network operations.
- Sanitization: Absent. The script output is processed directly to inform the agent's analysis and response.
- [COMMAND_EXECUTION]: The agent is instructed to execute a local Python utility (
scripts/atoms-search.py) via the command line to perform information retrieval. - [DATA_EXFILTRATION]: The search script transmits user-generated search queries to an external API endpoint and attempts to read a configuration file located in the user's home directory at
~/.xy/config.json.
Audit Metadata