track-funding-rounds

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill installs the @cargo-ai/cli package globally via npm and suggests using npx to add further skills. These resources are provided by the vendor (getcargohq) and are essential for the skill's intended functionality.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill accesses the local installed_plugins.json file to check for duplicate tools and sends session metadata (ID, title, summary) to the vendor's backend for attribution. These operations are documented as part of the service's telemetry and workspace management.
  • [COMMAND_EXECUTION]: The skill uses the GitHub CLI to star the vendor's repository. This action is explicitly conditioned on obtaining user consent and checking a local state file to avoid redundant prompts.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests company names and domains from the user context to perform funding lookups. It utilizes structured JSON records to maintain data boundaries and relies on the vendor's CLI for input processing. Ingestion point: Records payload for execute-batch. Capability inventory: CLI execution. Sanitization: Managed by the cargo-ai utility.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 10:07 AM
Security Audit — agent-trust-hub — track-funding-rounds