architecture-decision-records
Pass
Audited by Gen Agent Trust Hub on Jul 29, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon data stored in local documentation files.
- Ingestion points: The agent reads files from the
docs/adr/directory, including an index file and individual markdown ADR documents (SKILL.md). - Boundary markers: There are no explicit delimiters or system instructions defined to prevent the agent from obeying instructions that might be embedded within the content of the ADR files.
- Capability inventory: The skill is capable of reading from and writing to the local file system, specifically within the
docs/adr/directory (SKILL.md). - Sanitization: No sanitization, validation, or filtering is applied to the content read from ADR files before it is processed by the model to answer architectural questions.
Audit Metadata