ADR + Spec mandatory gate before any project or feature development. Architecture Decision Records and Software Design Docs enforce design clarity. Use when starting a new project, initiating non-trivial features, or discussing architecture decisions and design docs.
Use for continuing, resuming, or closing substantial multi-session work; chained AI sessions; blind-AI handoff; living docs; roadmap/spec/progress/session-handoff/CHANGELOG updates; next slash goal or next-session prompt; Git/PR/worktree closeout. NOT for trivial one-off edits or urgent one-command ops.
Use when planning a new project or feature, turning an idea or specification into an executable plan, designing architecture, establishing a roadmap or phase, or coordinating multi-session or multi-agent builds. Activate on "plan this build", "turn this into tasks", "make it a plan", or when tasking/swarm coordination…
Maintain configuration and capability parity across multiple AI agent runtimes (Claude, Codex, Cursor, Gemini, custom agents). Use when syncing rules, skills, MCPs, configs, or credentials across team AI tools, or when debugging drift between agents.
Validate before building — enforce product/market fit analysis, competitive positioning, and scope discipline before implementation. Use when starting a new product, MVP definition, feature scope discussion, competitive analysis, or any business-facing decision that precedes development.
Context preservation strategy for multi-session projects during context window compaction. Use when running /compact, when auto-compact is triggered, or when determining what state to preserve across session boundaries to maintain continuity.
Systematic debugging protocol for AI agents — pivot evidence-collection strategy after two failed reproduction rounds; treat available evidence (screenshots, console output, error messages) as ground truth from round one. Use when troubleshooting bugs, intermittent failures, error reproduction, or failed…
DNS lookup, WHOIS, passive-DNS history, IP reputation, and subdomain enumeration procedures. Auto-load when nslookup / dig / whois / host is mentioned, when investigating DNS records (A, NS, MX, CNAME, TXT), when researching historical nameservers, when checking IP reputation (AbuseIPDB, ipinfo.io), when enumerating…
Documentation accompanies code — no code is done without docs. Enforces a minimum documentation checklist per session, maintains a canonical documentation ledger across projects, and synchronizes documentation rules across all tools and agents. Use when writing code, creating features, modifying APIs, establishing new…
No intermediate options in trade-offs — every choice presented must be a path to excellence, never "accept broken state temporarily" or "good enough for now." Use when presenting trade-off decisions, discussing defect fixes, proposing feature scope, debating technical debt, MVP boundaries, and sequencing.
Always use Pull Requests — never merge locally into main. Workflow for coordinator sessions to batch-merge open PRs. Pre-work scan of open pull requests before starting feature/bug work. Use when working with git, mentioning merge, PR, pull request, gh pr, git push origin main, branch protection, squash merge, or code…
Hooks as enforcement mechanism for AI behavior — use when a written rule fails repeatedly in generation flow and deterministic structural enforcement is needed. Triggers include detecting cyclic rule failure, asking about Claude Code hooks/lifecycle events (SessionStart, UserPromptSubmit, PreToolUse, PostToolUse…
Enforce hook lifecycle synchronization across all AI agent tools. When creating, modifying, or removing hooks in any AI tool (Claude Code, Codex CLI, Cursor, Gemini, Kimi, etc.), propagate changes to all equivalent tools in the same session. Use when implementing cross-tool runtime enforcement.
Infrastructure-as-Code from the first commit — every new project must be rebuildable from git clone + one command, with declarative state in code and idempotent execution. Use when scaffolding new projects, bootstrapping infrastructure, choosing IaC tools (Ansible/Terraform/SST/Fastlane), or setting up infra/…
Measure before building — enforce impact metrics definition before project/feature development. Use when starting projects, defining features, or setting success criteria.
Verify large payload fetches are COMPLETE — anti silent-truncation. Cross-check size vs expected, follow all pagination cursors. Use when fetching from paginated APIs, large databases, HAR files, log dumps, transcripts, or any payload >50k chars.
Automatic documentation updates at end of substantive sessions. Update CHANGELOG + session-handoff without being asked. Triggers on session-end signals (farewells, wrap-up requests, topic pivots after coherent work, or context limits approaching).
Select the right LLM for a task by consulting benchmarks and real performance data. Use when choosing between models (chatbot, agent, text classification, code generation, etc.), comparing model performance, or validating model availability and pricing.
Auto-start offline local services — detect connection errors, daemon failures, and start services programmatically instead of reporting them as blockers. Use when local service is unreachable, daemon not running, connection refused, port not listening, service offline.
Manage multi-phase projects using separate sessions per phase instead of compact, with durable handoff files. Use when work spans multiple sessions, phases, handoffs, or when context degradation becomes a concern.
Project bootstrap standard — guarantees all 25 core engineering and product disciplines are consulted in lifecycle order. Use when starting a new project, repository, feature, or phase. Prevents auto-trigger gaps by making the standard enforceable through a checklist.
Use Notion REST API (not MCP) for batch operations (≥3 rows). Applies to database queries, batch updates, scans, pagination, file uploads, database creation, and page replication. MCP is acceptable only for single-page CRUD operations.