ecdash-api

Fail

Audited by Gen Agent Trust Hub on Jun 19, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill accesses a sensitive authentication token stored at /root/.secrets/ecdash_token to authenticate with its dashboard API.
  • [DATA_EXFILTRATION]: The contents of the sensitive credential file are read and transmitted via an Authorization header in network requests to an external service hosted on Railway.app.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands to manage local files, perform network requests via curl, and manage Git repository states including commits and pushes.
  • [REMOTE_CODE_EXECUTION]: Utilizes Python's importlib.util to dynamically load and execute functions from a local script (scripts/update-dashboard.py), which is a technique that can be leveraged to run arbitrary code if the source file is compromised.
  • [EXTERNAL_DOWNLOADS]: Communicates with an external dashboard API at jay-portfolio-production.up.railway.app to synchronize task data and update project statuses.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 19, 2026, 06:52 PM
Security Audit — agent-trust-hub — ecdash-api