read-slack-messages

Pass

Audited by Gen Agent Trust Hub on Jun 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The documentation includes instructions to install the Bun runtime from the official bun.sh domain. This is a well-known service and the installation is a prerequisite for executing the skill scripts.
  • [COMMAND_EXECUTION]: The script uses standard system commands to launch a web browser during the Slack OAuth login process. The commands are specific to the operating system (open, start, or xdg-open) and the URLs are securely constructed from configuration values.
  • [DATA_EXFILTRATION]: The skill interacts with the official Slack API to fetch message, user, and file data. All network traffic is restricted to slack.com domains.
  • [CREDENTIALS_UNSAFE]: The skill manages sensitive Slack access tokens. It implements security hardening by storing tokens in a gitignored directory with restricted file permissions (0600) and includes warnings against printing secrets to the logs or chat.
  • [SAFE]: The skill processes untrusted Slack messages, which constitutes an indirect prompt injection surface. However, the risk is mitigated as the skill's capabilities are restricted to Slack-specific operations and local file downloads to defined paths. Ingestion point: scripts/slack.ts (Slack API responses). Boundary markers: Absent. Capability inventory: File writing (tokens and downloads), network access to Slack, browser execution. Sanitization: Message text is cleaned of Slack-specific formatting characters.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 14, 2026, 03:44 AM
Security Audit — agent-trust-hub — read-slack-messages