night-market-operations

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process data from the local repository which creates an attack surface for data-driven instruction overrides. * Ingestion points: Repository manifests and plugin source files are read during commands like 'make test', 'make validate-all', and 'scripts/clawhub_export.py'. * Boundary markers: There are no explicit delimiters or 'ignore' instructions provided to isolate the ingested repository content from the agent's instructions. * Capability inventory: The skill has the ability to modify the filesystem through automated linting fixes and version-bumping scripts. * Sanitization: The runbook does not mention any validation or sanitization steps for the data ingested from the repository.
  • [DYNAMIC_EXECUTION]: The skill relies on the execution of a pre-compiled binary ('skrills') and several local scripts (e.g., 'update_versions.py') at runtime to automate maintenance tasks.
  • [COMMAND_EXECUTION]: The skill provides procedures for executing complex administrative commands via 'make' and 'uv', which include filesystem mutations and participation in the CI/CD release pipeline.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 10:05 AM
Security Audit — agent-trust-hub — night-market-operations