10 docs tagged with "agent-ops"

Index of runbooks for Claude Code agents — task startup, git safety, knowledge maintenance, QA, deployment checklist, feature registry, and decision logging

Purpose: Tracks the classification and lifecycle status of all documentation across the ChurchWiseAI portfolio. Agents consult this to know which docs to trust.

Git safety protocol for concurrent Claude agents — always create your own branch, never reuse another agent's branch, rebase before pushing, and never force-push to main/master

Reference for how Claude Code agents are configured — session-start hook, available MCP tools, skills (slash commands), CLI access, and how multiple agents coordinate safely

How to self-review implementation as a senior QA engineer — using QA_CHECKLIST.md and AGENT_QUALITY_PRINCIPLES.md together to find critical, important, and minor issues before shipping

Pre-work checklist before writing any code — read CLAUDE.md, check the feature registry, verify knowledge/ docs, create a branch, and confirm task scope with the founder

When and how to append to DECISION_LOG.md — the shared institutional memory that prevents agents from re-litigating past decisions or repeating mistakes across sessions

How to check FEATURE_REGISTRY.md before building anything new — the decision tree for approved vs unapproved features, cross-property ownership rules, and how to update the registry when adding features

The rule for keeping knowledge/ docs in sync with code changes — which code files to monitor, how to run pnpm derive, and why stale knowledge is a real-time product failure

Every item an agent must verify before merging to a deploy branch — build passes, env vars set, QA checklist complete, no hardcoded test data, and marketing copy verified