childhood
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to read and evaluate external data, including issue reports, merged-PR comments, and review threads. These external sources could contain malicious instructions designed to influence the agent's behavior during the observation and repair phase.
- Ingestion points: Issue reports, PR comments, and reviews (SKILL.md).
- Boundary markers: Absent; the skill does not specify the use of delimiters or instructions to ignore embedded commands within external data.
- Capability inventory: Code modification ("focused fixes"), command execution ("run a small installed CLI smoke"), and project state management (
work_state_write) (SKILL.md). - Sanitization: No explicit sanitization or validation of external content is defined before the content is verified against the codebase.
- [COMMAND_EXECUTION]: The agent is instructed to run a "CLI smoke" test from an isolated directory after a release. While this is a standard verification practice, it involves executing binaries or scripts that were recently merged or published.
- [EXTERNAL_DOWNLOADS]: The skill involves downloading published artifacts from registries to verify hashes and entry points. These operations are performed on artifacts that the agent is specifically tasked with monitoring and verifying.
Audit Metadata