pp-agoda

Pass

Audited by Gen Agent Trust Hub on Aug 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches the agoda-pp-cli and agoda-pp-mcp tools from the vendor's GitHub repository (github.com/mvanhorn/printing-press-library) via go install.
  • [EXTERNAL_DOWNLOADS]: Installs the CLI utility via the vendor's NPM package (@mvanhorn/printing-press-library) using npx.
  • [COMMAND_EXECUTION]: The skill operates by executing shell commands using the agoda-pp-cli binary to perform hotel searches, rank properties, and manage local state.
  • [DYNAMIC_EXECUTION]: Utilizes a 'Playbook' system that records command sequences and replays them for future queries. This involves dynamic command synthesis from a local session journal based on successful past executions.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from Agoda, such as hotel reviews and property descriptions, which are interpolated into the agent's context.
  • Ingestion points: Data enters the context through agoda-pp-cli reviews and agoda-pp-cli hotels search outputs.
  • Boundary markers: None explicitly defined in the skill instructions to separate external data from agent instructions.
  • Capability inventory: The skill has access to shell execution via the CLI and network operations to Agoda domains.
  • Sanitization: No specific sanitization or filtering logic for external hotel data is described.
  • [CREDENTIALS_UNSAFE]: Manages authenticated sessions via AGODA_COOKIE environment variables and stores local data in a credentials.toml file located in the user's data directory.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 28, 2026, 03:26 PM
Security Audit — agent-trust-hub — pp-agoda