test-evidence-review
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_CONTEXT_INJECTION
Full Analysis
- [DYNAMIC_CONTEXT_INJECTION]: The skill uses the exclamation-backtick syntax to execute a shell command at load time for configuration resolution. Evidence: !bash "${CLAUDE_SKILL_DIR}/../../hooks/yaml-helper.sh" resolve_config --keys automation,workflow,qa.level,testing.strict in the skill body.
- [COMMAND_EXECUTION]: The skill executes local project-specific bash scripts to verify evidence timestamps. Evidence: Shell execution of .claude/scripts/review-receipts.sh in Section 5.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted data from story files and evidence documents, creating a surface for indirect prompt injection. 1. Ingestion points: Reads story files from production/epics/ and evidence documents from production/qa/evidence/. 2. Boundary markers: The skill does not implement explicit boundary markers or delimiters to separate untrusted content from the agent's internal instructions. 3. Capability inventory: The skill has access to file reading/writing tools (Read, Write) and shell command execution (Bash). 4. Sanitization: No evidence of sanitization or filtering of external content before processing is present in the instructions.
Audit Metadata