reconcile
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and process external content, including source code, developer comments, and vendor API/SDK schemas, to update a documentation graph. This creates a potential surface for indirect prompt injection if those external sources contain malicious instructions.
- Ingestion points: Source code files, code comments (комментарий-портянка), and external surfaces such as vendor APIs and schemas.
- Boundary markers: None defined in the instructions to prevent the agent from following instructions embedded in the code or comments.
- Capability inventory: The skill utilizes vendor tools (
iskron_orient) and involves writing/weaving operations to modify graph nodes. - Sanitization: No sanitization or validation steps are described for the external content before it is used to update the graph.
- [SAFE]: The skill uses vendor-specific tools (
iskron_orient) and terminology consistent with the author's ecosystem to manage project context and documentation. - [SAFE]: No hardcoded credentials, malicious remote code execution patterns, or persistence mechanisms were detected.
Audit Metadata