The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes standard command-line utilities including curl, gh, git, and vercel to retrieve status information, deployment logs, and health check results. These commands are used solely for diagnostic data collection.
[EXTERNAL_DOWNLOADS]: The skill fetches health check status and incident summaries from Superset-owned infrastructure at relay.superset.sh, realtime.superset.sh, api.superset.sh, and status.superset.sh. These requests are directed to the vendor's own domains for the purpose of monitoring service availability.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources such as Sentry issue titles, Git commit messages, and Vercel logs. While these sources are untrusted, the skill is restricted to read-only operations and instructs the agent to produce a draft for human approval, providing a necessary manual verification step before the output is used.

incident-triage