skills/fastcrw/crw/crw-search/Gen Agent Trust Hub

crw-search

Pass

Audited by Gen Agent Trust Hub on Oct 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes data from external web sources (search results, titles, and snippets) which are then presented to the agent. This content is untrusted and could contain instructions designed to influence or hijack the agent's behavior.
  • Ingestion points: Results from the crw search command or REST API calls are piped into the agent's context.
  • Boundary markers: The instructions do not specify delimiters or "ignore instructions" warnings for the ingested web content.
  • Capability inventory: The skill can execute shell commands via Bash(crw:*) and Bash(curl:*), and has file Read permissions.
  • Sanitization: There is no evidence of sanitization or filtering of the retrieved snippets before they are processed by the agent.
  • [COMMAND_EXECUTION]: The skill relies on the execution of the crw CLI binary and curl to perform search operations. This behavior is consistent with the skill's stated purpose of providing a command-line search interface and interacting with its own self-hosted backend API.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 12, 2026, 09:36 AM