skills/necmttn/osome-skill/osome/Gen Agent Trust Hub

osome

Pass

Audited by Gen Agent Trust Hub on May 11, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill's implementation aligns with its stated purpose of automating the Osome document upload workflow. Authentication is handled via session cookies provided by the user in an environment file, which is a common and acceptable practice for automation tools targeting internal or reverse-engineered APIs.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests external data (transaction descriptions) from the Osome API which is subsequently processed in CLI outputs and PDF templates.
  • Ingestion points: Transaction metadata is fetched from https://my.osome.com/api/v2 in the scripts/osome.ts and lib/api/osome.ts files.
  • Boundary markers: The skill does not use specific delimiters or instructions to isolate the ingested transaction data from agent instructions.
  • Capability inventory: The skill's capabilities include performing network requests via fetch (limited to the Osome API and generated S3 URLs), reading local files, and executing a headless browser session via Puppeteer.
  • Sanitization: The skill performs standard JSON parsing of the API responses but does not include explicit sanitization or filtering of the natural language fields within those responses.
Audit Metadata
Risk Level
SAFE
Analyzed
May 11, 2026, 12:39 PM
Security Audit — agent-trust-hub — osome