xy-ai-workflow
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a local Python script,
scripts/atoms-search.py, to retrieve methodology and evidence (referred to as 'atoms') from a knowledge base. The agent is instructed to execute this script to find relevant case studies and rules for the auditing process.\n- [EXTERNAL_DOWNLOADS]: Theatoms-search.pyscript includes functionality to perform HTTPS POST requests tohttps://api.xyskill.xyz/v1/atoms/search. This allows the skill to fetch knowledge base updates from the vendor's remote API when local data is insufficient. This domain is managed by the skill's author (xyaz1313).\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a potential surface for indirect prompt injection by ingesting data from external files and APIs.\n - Ingestion points: Data enters the context via the local
references/atoms.jsonlfile and the remote vendor API.\n - Boundary markers: The skill does not employ specific technical delimiters or 'ignore embedded instructions' warnings for the retrieved 'atom' content, relying instead on citation instructions.\n
- Capability inventory: The skill has the ability to execute the local search script and read files within its own directory structure.\n
- Sanitization: No specific filtering or sanitization of the retrieved knowledge content for prompt injection payloads was identified in the script logic.
Audit Metadata