adr
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute
git config user.nameto identify the author for new documentation records. This is a standard developmental procedure for attribution. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from existing project files to maintain an index and link related documentation, creating a surface where malicious instructions in those files could theoretically influence agent behavior.
- Ingestion points: The agent reads
docs/adr/INDEX.md, individual ADR files (e.g.,ADR-001-slug.md), anddocs/fdr/INDEX.md(if it exists) to perform indexing and cross-referencing. - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious content within the files being read.
- Capability inventory: The skill performs local file system read and write operations and executes a shell command to retrieve git configuration.
- Sanitization: The skill does not implement specific sanitization or validation logic for the content retrieved from the documentation files.
Audit Metadata