book-hotel

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from preceding search tasks to execute high-capability financial commands.
  • Ingestion points: Processes package and session identifiers from search results in SKILL.md.
  • Boundary markers: Present; the instructions explicitly require the agent to summarize and obtain manual user confirmation of all booking details and prices before proceeding.
  • Capability inventory: Wallet balance verification, on-chain payment execution, and booking status recovery via shell commands.
  • Sanitization: Absent; the skill relies on user confirmation rather than automated input validation.
  • [COMMAND_EXECUTION]: Executes shell commands to interface with crypto wallets and booking APIs. Evidence: Usage of npx awal@latest for payments and npx @tvl-justin/travel-cli@latest for status checks.
  • [EXTERNAL_DOWNLOADS]: Downloads and runs external tools from the NPM registry at runtime. Evidence: Fetches awal@latest (associated with Coinbase) and @tvl-justin/travel-cli@latest (associated with the author).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 05:57 PM
Security Audit — agent-trust-hub — book-hotel