genfeed-connector

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The detect function in lib/detect.ts executes the genfeed status command using execFileSync to verify if the Genfeed CLI tool is installed and authenticated on the host system. This is a standard capability detection pattern for vendor-specific tooling.
  • [DATA_EXPOSURE]: The skill reads API keys (defaulting to GENFEED_API_KEY) and platform-specific access tokens (e.g., X_BEARER_TOKEN, LINKEDIN_ACCESS_TOKEN) from environment variables. These credentials are used to authenticate requests to the Genfeed API or are resolved for other worker skills. The skill explicitly avoids writing these secret values to the local filesystem, storing only the names of the environment variables in its configuration files.
  • [EXTERNAL_DOWNLOADS]: In its connected mode, the skill performs network requests to api.genfeed.ai using the global fetch API. These requests are used to synchronize content manifests and obtain short-lived, platform-scoped tokens from the Genfeed vault. These operations are directed at the official infrastructure of the skill's author.
  • [INDIRECT_PROMPT_INJECTION]: The CLI interface gf.ts ingests JSON data from stdin to create content items or record metrics. While this represents a surface for ingesting untrusted data from other tools or web sources, the skill treats this data as passive state and does not execute instructions contained within the manifest fields.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:43 PM
Security Audit — agent-trust-hub — genfeed-connector