senpi-qa

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill implements a secure isolation strategy by instructing the agent to use a dedicated, temporary directory for testing (SENPI_CODING_AGENT_DIR), explicitly avoiding modifications to the default ~/.senpi/agent directory.
  • [SAFE]: The provided resolve-evidence-dir.mjs script uses strict regular expression validation (/^[a-z0-9]+(?:-[a-z0-9]+)*$/) for user-provided slugs, effectively preventing directory traversal attacks and ensuring that artifacts are stored within the repository's sanctioned evidence path.
  • [COMMAND_EXECUTION]: The skill coordinates the execution of various local test scripts and binaries (node, bun, senpi) which is the primary intended function of a QA tool. These operations are scoped to the repository's packages and do not involve remote code execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 12:07 PM
Security Audit — agent-trust-hub — senpi-qa