platform

Pass

Audited by Gen Agent Trust Hub on Mar 22, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions to install external dependencies from the npm registry, specifically '@effect/platform' and its platform-specific variants '@effect/platform-node' and '@effect/platform-bun'.
  • [COMMAND_EXECUTION]: The skill documents and enables powerful system capabilities through the 'FileSystem' and 'Terminal' modules, which allow for reading, writing, and deleting files, as well as performing terminal interactions.
  • [DATA_EXFILTRATION]: The skill's primary functionality involves combining file system access with network capabilities ('HttpClient'), which technically provides a functional mechanism for transferring local data to remote endpoints.
  • [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by processing untrusted external data.
  • Ingestion points: Network response bodies and server request parameters/bodies retrieved via '@effect/platform' (SKILL.md).
  • Boundary markers: Absent. No explicit delimiters or specific warnings to ignore embedded instructions are demonstrated in the example code snippets.
  • Capability inventory: File system operations (read/write/delete), terminal interaction, and outbound network requests (SKILL.md).
  • Sanitization: The skill documents best practices including 'Schema validation' using the 'effect' schema library to validate and sanitize the structure of external data.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 22, 2026, 01:53 PM