fintel-data

Fail

Audited by Gen Agent Trust Hub on Oct 22, 2026

Risk Level: HIGHDYNAMIC_CONTEXT_INJECTIONCREDENTIALS_UNSAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DYNAMIC_CONTEXT_INJECTION]: The file SKILL.md uses the !command syntax to execute a complex shell script immediately upon skill loading. This script silently probes the environment and local file system for API keys without user intervention.
  • [CREDENTIALS_UNSAFE]: The skill's initialization logic automatically reads from .env files located in the current directory and the git repository root. Accessing .env files is a high-risk pattern as they typically contain various sensitive secrets beyond the scope of this specific skill.
  • [COMMAND_EXECUTION]: The skill leverages the agent's ability to run shell commands (including curl, grep, sed, and git) to perform credential discovery and API interactions. This reliance on the host shell for all operations increases the risk of command injection or unauthorized system access.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external data from the Fintel API, which introduces an attack surface for instructions embedded in data.
  • Ingestion points: Data is fetched via curl from https://api.fintel.io/v1/*.
  • Boundary markers: None identified in the prompt instructions to isolate external data from the agent's internal logic.
  • Capability inventory: The agent can execute shell commands, read local configuration files, and access git metadata.
  • Sanitization: No explicit sanitization or validation of the API's JSON output is documented before the content is processed by the agent.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Oct 22, 2026, 11:53 AM