doc-adr-reviewer
Warn
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill explicitly instructs the agent to execute bash commands (
sha256sum <file_path> | cut -d' ' -f1andgrep) for hash calculation and cache verification. These commands use file paths extracted dynamically from document content. - [PROMPT_INJECTION]: This skill is vulnerable to Indirect Prompt Injection (Category 8).
- Ingestion points: ADR files in
docs/05_ADR/, BDD feature files in../04_BDD/, and the.drift_cache.jsonfile. - Boundary markers: None identified in the prompt instructions to isolate untrusted content from the commands.
- Capability inventory: The skill has the capability to execute shell commands, read/write local files (reports, cache), and calculate cryptographic hashes.
- Sanitization: There is no mention of path sanitization or validation for the extracted upstream references before they are passed to the bash shell, which could lead to command injection via malicious filenames or path traversal.
Audit Metadata