Skills
Navigate
skills/prefactordev/typescript-sdk/inspect-agent-run-with-prefactor-cli/Gen Agent Trust Hub

inspect-agent-run-with-prefactor-cli

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the prefactor CLI tool to run diagnostic commands including agent_instances agent_context and agent_spans list to retrieve runtime telemetry.
  • [CREDENTIALS_UNSAFE]: The skill requires a PREFACTOR_API_TOKEN to be provided by the user to authenticate the prefactor CLI with the remote telemetry service.
  • [PROMPT_INJECTION]: The skill processes external trace data and codebase files, which constitutes an indirect prompt injection surface.
  • Ingestion points: Trace and context data downloaded via the CLI and existing source code files from the local repository.
  • Boundary markers: None specified; instructions do not provide delimiters to separate untrusted trace content from agent instructions.
  • Capability inventory: Shell command execution via the prefactor CLI and file system read access to the local codebase.
  • Sanitization: No explicit validation or filtering of the incoming trace payloads is mentioned.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 04:22 AM
Security Audit — agent-trust-hub — inspect-agent-run-with-prefactor-cli