kelly-portrait-retouch

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The script scripts/retouch.mjs executes a local Swift script scripts/detect-faces.swift using the swift interpreter via node:child_process.execFile. This is used for face detection on macOS platforms.
  • [DATA_EXFILTRATION]: The scripts scripts/sync-candidate.mjs and scripts/lib/portrait-busabase.mjs perform network operations to upload local image files (source, output, and comparison) to a remote Busabase instance configured in the environment.
  • [CREDENTIALS_UNSAFE]: The skill's synchronization logic in scripts/lib/portrait-busabase.mjs reads sensitive configuration including BUSABASE_API_KEY from environment variables to facilitate authentication with the Busabase API.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted local image files which constitutes an indirect prompt injection attack surface.
  • Ingestion points: The INPUT image path processed by scripts/retouch.mjs.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded content are used when processing image data or results.
  • Capability inventory: File system access via sharp and fs, local command execution of swift, and network exfiltration to Busabase via fetch.
  • Sanitization: While image processing is performed by the established sharp library, metadata and counts extracted from the images are returned directly to the agent's context.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 07:29 AM