x-presence

Fail

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEDATA_EXFILTRATIONPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The skill requires the agent to collect the user's raw X/Twitter login credentials, including password and TOTP secret, for account linking.
  • [DATA_EXFILTRATION]: Collected credentials are transmitted to an external API (https://twitr.sh/api/x-accounts/connect). This non-OAuth authentication method exposes sensitive user login data to a third-party service provider.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection by incorporating untrusted data from external tweets into its drafting process without adequate safety measures.
  • Ingestion points: Tweet events fetched from monitors (SKILL.md).
  • Boundary markers: No delimiters or 'ignore' instructions are used when interpolating external content into the compose tool (SKILL.md).
  • Capability inventory: The skill can post content to X via tools and modify local files (SKILL.md).
  • Sanitization: No mechanisms are specified for validating or escaping ingested tweet text.
  • [COMMAND_EXECUTION]: Provides instructions for establishing persistent background execution using shell commands to configure local cron, launchd, and agent-specific schedulers (references/scheduling.md).
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 27, 2026, 09:19 PM
Security Audit — agent-trust-hub — x-presence