Skills
Navigate
skills/reticlehq/reticle/fix-what-i-pointed-at/Gen Agent Trust Hub

fix-what-i-pointed-at

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided notes and element labels from the Reticle application, which introduces a surface for indirect instructions.
  • Ingestion points: Data returned from the reticle_session tool, specifically the note and label fields within the marks array.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between data and commands within these notes.
  • Capability inventory: The agent has access to file modification, application interaction via reticle_act_and_wait, and session management tools.
  • Sanitization: No sanitization or validation of the input strings is mentioned.
  • [COMMAND_EXECUTION]: The skill references an initialization command for setting up the environment.
  • Evidence: Mentions RETICLE_INSTALL_SOURCE=npx_skill npx @reticlehq/server@latest init for setting up the Reticle server.
  • [EXTERNAL_DOWNLOADS]: The skill references external vendor resources for documentation and package initialization.
  • Evidence: References npx @reticlehq/server@latest and curl https://docs.reticle.sh/llms.txt for additional information.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 09:20 AM
Security Audit — agent-trust-hub — fix-what-i-pointed-at