DenchClaw

Connected app integration recipes for Dench Integrations (Gmail, Slack, GitHub, Notion, Google Calendar, Linear, Stripe, YouTube, and 500+ more)

Build data-driven DenchClaw apps with full CRUD access to workspace objects (.object.yaml tables), DuckDB queries and mutations, data dashboards with Chart.js and D3.js, and interactive tools.

Build and manage DenchClaw apps — self-contained web applications that run inside the workspace with access to DuckDB data, workspace objects, AI chat, and the full DenchClaw platform API.

DuckDB schema initialization, field types reference, auto-generated PIVOT views, and SQL CRUD operations for workspace objects, fields, and entries.

Full 3-step workflow for creating workspace objects (SQL → filesystem → verify), CRM patterns for common object types, kanban boards, and the post-mutation checklist.

Manage DuckDB CRM data, aggressive relation-linked fields, and synced markdown documents in the workspace. Use when creating or updating objects, fields, entries, foreign-table links, row notes, or entry-linked edit logs.

.object.yaml format and template, view type settings (kanban, calendar, timeline, gallery, list), saved views with filter operators, and date format rules.

Build AI-powered DenchClaw apps that interact with the OpenClaw agent — create chat sessions, send and receive messages with streaming, expose app tools for agent invocation, and access agent memory.

Build 2D and 3D games as DenchClaw apps using p5.js, Three.js, Matter.js, and other game libraries. Covers game architecture, sprites, physics, particles, audio, tilemaps, and complete game examples.

Platform API reference for DenchClaw apps — UI integration, per-app storage, HTTP proxy, real-time events, inter-app messaging, cron scheduling, webhooks, clipboard, context, and widget mode.

Connected app tool recipes for Dench Integrations (Gmail, Slack, GitHub, Notion, Google Calendar, Linear, Stripe)

Document management with markdown files, always-on entry documents linked through the documents table, and mutation/edit logs for CRM entries. Use when creating or updating row notes, entry notes, detail pages, or document-linked CRM content.

Report generation with .report.json format, interactive chart types (bar, line, area, pie, donut, radar, scatter, funnel), panel sizes, filter types, inline chat reports, and post-report checklist.

Performance benchmarking. Baseline page load times, Core Web Vitals, and resource sizes. Compare before/after on every PR. Catch bundle size regressions before they ship.

Safety guardrails — warns before destructive commands. Activate by saying "be careful" or when working with production systems. Warns on: rm -rf, DROP TABLE, force-push, git reset --hard, DELETE without WHERE.

Build a complete design system from scratch. Research the landscape, propose safe choices AND creative risks, generate DESIGN.md. Use when creating a new design system or establishing design foundations for a project.

Design audit of existing code. Same methodology as plan-design-review, but applies fixes directly with atomic commits and before/after evidence. Use when asked to audit the visual design of implemented code.

Update all project documentation to match what was just shipped. Catches stale READMEs, outdated architecture docs, missing changelog entries. Use after shipping features or when docs drift from code.

Restrict file edits to one directory. Prevents accidental changes outside scope while debugging or working on a focused task. Use when you want to scope edits to a specific module or directory.

Systematic root-cause debugging. Four phases: investigate, analyze, hypothesize, implement. Iron Law: no fixes without root cause investigation first. Use when debugging errors, unexpected behavior, or troubleshooting.

Merge the PR, wait for CI and deploy, verify production health. Takes over after ship. One command from "approved" to "verified in production."

YC Office Hours — forcing questions that reframe the product before code. Two modes: Startup mode (demand reality, status quo, narrowest wedge) and Builder mode (design thinking for side projects and hackathons). Saves a design doc that feeds into plan-ceo-review and plan-eng-review. Use when asked to brainstorm, explore an idea, or before any planning skill.

CEO/founder-mode plan review. Rethink the problem, find the 10-star product, challenge premises. Four modes: SCOPE EXPANSION, SELECTIVE EXPANSION, HOLD SCOPE, SCOPE REDUCTION. Use when reviewing strategy, questioning scope, or before engineering review.

Design review of plans before implementation. Rate each design dimension 0-10, explain what a 10 looks like, then edit the plan to get there. AI Slop detection. Use when reviewing UI/UX plans or design decisions.

Engineering architecture review. Lock in data flow, diagrams, edge cases, test strategy, and failure modes. Forces hidden assumptions into the open. Use when reviewing technical plans, architecture decisions, or before implementation.

Systematically QA test a web application and fix bugs found. Three tiers: Quick (critical/high only), Standard (+ medium), Exhaustive (+ cosmetic). Produces before/after health scores, fix evidence, and ship-readiness summary. Use when testing a feature, finding bugs, or verifying deployments.

Pre-landing PR review. Analyzes diff for SQL safety, race conditions, LLM trust boundary violations, conditional side effects, and structural issues. Auto-fixes obvious issues, asks about complex ones. Use when reviewing code before merge, or when asked for "code review".

Release engineering — sync main, run tests, audit coverage, push branch, open PR. Bootstraps test frameworks if needed. One command from "code complete" to "PR ready for review".

Structured AI-assisted development workflow with specialist roles — Think, Plan, Build, Review, Test, Ship, Reflect. Adapted from garrytan/gstack (MIT).

Post-deploy monitoring loop. Watches for console errors, performance regressions, and page failures. Periodic screenshots and anomaly detection. Use after deploying to production.

Maximum safety mode — combines careful (destructive command warnings) with freeze (edit scope restriction). Use when working with production systems or making sensitive changes.

Same QA methodology as the qa skill but report-only — no code changes. Use when you want a pure bug report without fixes, or when reviewing someone else's work.

Weekly retrospective with developer stats, shipping streaks, test health trends, and growth opportunities. Use for weekly check-ins or when asked for a retrospective.

Remove the freeze/guard edit boundary. Restores unrestricted file editing.