skills/reelary/skills/find-footage/Gen Agent Trust Hub

find-footage

Pass

Audited by Gen Agent Trust Hub on Oct 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to use npx -y reelary, which downloads and executes the Reelary package from the official npm registry. This is the primary intended functionality provided by the vendor.
  • [COMMAND_EXECUTION]: The skill provides CLI command templates for searching and opening files (e.g., npx -y reelary search "<query>" --json). These commands are executed locally to facilitate the search functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided search queries to find specific footage.
  • Ingestion points: User search queries provided as natural language strings (e.g., "find the clip where").
  • Boundary markers: None explicitly defined in the instructions for user input; the skill relies on the agent's default command-line argument handling.
  • Capability inventory: File search (local metadata only), listing recognized individuals/places, and opening files in the local Reelary application.
  • Sanitization: The skill does not provide explicit sanitization instructions, relying on the agent's underlying execution environment to handle command-line argument safety.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 15, 2026, 08:20 AM