test
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from project files (e.g., AGENTS.md, CLAUDE.md, README) and ticket systems, creating an attack surface for indirect prompt injection.
- Ingestion points: The agent reads command instructions and acceptance criteria from external repository files and user-provided ticket text.
- Boundary markers: No specific delimiters or safety warnings are provided to the agent to ignore instructions embedded within the ingested repository data.
- Capability inventory: The skill possesses extensive capabilities including arbitrary command execution (terminal), file system modifications (writing tests), network access (HTTP requests), and web automation (browser use).
- Sanitization: There is no evidence of sanitization or validation of the external content before it influences the agent's behavior.
- [COMMAND_EXECUTION]: The skill explicitly directs the agent to discover and run shell commands found in documentation or project scripts (e.g., lint, format, and test commands), which may lead to execution of malicious commands if the source files are compromised.
- [DYNAMIC_EXECUTION]: The instructions require the agent to dynamically author and execute new test scripts when existing coverage is insufficient, representing a form of runtime code generation and execution.
Audit Metadata