konecty-meta-document

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFECREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The script 'scripts/meta_document.py' reads authentication and host configuration from local user files to retrieve API access tokens.\n
  • Evidence: Accesses '/.konecty/credentials' and '/.konecty/.env' to load 'KONECTY_TOKEN' and 'KONECTY_URL'.\n- [EXTERNAL_DOWNLOADS]: The skill communicates with external API endpoints to manage document metadata using the standard library.\n
  • Evidence: Performs GET and PUT requests to the configured Konecty server using 'urllib.request' for schema management endpoints under '/api/admin/meta'.\n- [PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface through its metadata retrieval functions.\n
  • Ingestion points: Fetches document schemas (including field labels and descriptions) from the remote server via the 'show' and 'fields' commands.\n
  • Boundary markers: Lacks explicit delimiters or 'ignore' instructions when displaying retrieved metadata to the agent context.\n
  • Capability inventory: Possesses administrative capabilities to modify document definitions, fields, and events on the server.\n
  • Sanitization: Metadata retrieved from the server is not filtered or escaped before being output to the console.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 07:26 PM