Skills
Navigate
skills/orcaqubits/agentic-commerce-skills-plugins/webmcp-setup/Gen Agent Trust Hub

webmcp-setup

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to fetch and process data from external websites and search results, creating a surface for indirect prompt injection.
  • Ingestion points: The skill uses WebFetch to retrieve W3C draft reports from webmachinelearning.github.io and Chrome implementation notes from developer.chrome.com. It also utilizes WebSearch for polyfill documentation and Chrome flag status.
  • Boundary markers: The instructions do not provide delimiters or specific directions to ignore instructions potentially embedded within the fetched external data.
  • Capability inventory: The skill utilizes Bash, Write, and Edit tools to scaffold the project, which could be misused if malicious instructions are ingested from external sources.
  • Sanitization: There is no evidence of sanitization or validation logic for content fetched via the network.
  • [EXTERNAL_DOWNLOADS]: The skill references and fetches documentation from external repositories and websites to guide the setup process.
  • Evidence: Mentions fetching content from webmachinelearning.github.io, developer.chrome.com, and chromestatus.com.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 06:33 PM