The Agent Skills Directory

[Operational Data Processing]: The skill instructs the agent to ingest and analyze external data sources, specifically production logs and code diffs. This is a necessary component of incident triage, though it creates a surface area for indirect prompt injection if those logs or diffs contain malicious text intended to influence the agent. In the context of SRE tasks, this is an expected behavior and the skill includes specific instructions to focus on timestamps and known issues rather than open-ended 'fishing'.
[Credential and Secret Handling]: The runbook asks for a commit SHA as a root cause identifier and references service names like 'stripe-api' or 'db-primary'. These are standard identifiers and do not involve hardcoded credentials, sensitive tokens, or unsafe secret management practices.

incident-triage-runbook