validate

Pass

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill is designed to run local CLI commands such as tsc, lint, and curl, along with existing repository scripts. These commands are used to gather live evidence for validating code health and functionality. The instructions specify running only narrow, purposeful commands to fill gaps in evidence, which aligns with standard developer workflows.
  • [PROMPT_INJECTION]: The skill ingests acceptance criteria from external sources including ACCEPTANCE.md, GOAL.md, and user-pasted text. This presents an indirect prompt injection surface where malicious instructions could be embedded in the processed data.
  • Ingestion points: Acceptance criteria are pulled from ACCEPTANCE.md, GOAL.md, tracker tickets, and user pastes.
  • Boundary markers: The skill does not define specific delimiters to isolate external data from its own instructions.
  • Capability inventory: The agent has access to shell command execution (via CLI tools) and browser-based testing tools.
  • Sanitization: No explicit sanitization or validation of the ingested criteria is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 27, 2026, 03:01 PM
Security Audit — agent-trust-hub — validate