Handbook
Agents and Cursor
Repo affordances for automation and IDE integration:
Updated
- For agents — map of safe tasks and verification commands.
- Safe edit matrix — risk tiers and required checks by area.
- Cursor — quick guide — open repo, install, verification matrix.
- Cursor integration — rules, skills, workflows under
.cursor/. - MCP sidecar — overview · MCP-SIDECAR (full) — LCDL as MCP tools in the IDE.
- MCP client — Python calling Playwright MCP and policies.
- Task authoring workflow
- Docs refactor workflow
- Reusable agent prompts — copy-paste prompt blocks.
Handbook map (stable entry URLs)
| Area | Purpose |
|---|---|
| Start | Install, adoption, troubleshooting |
| Tutorials — 101 · 201 · 301 | Progressive labs |
| Guides | Playwright discovery, task packs, graph |
| Reference | Client API, ContractSpec, schemas, catalog |
| Enterprise | Threat model, operations, compatibility |
| Contribute | Docs authoring, release checklist, website publishing |
Agents hub handbook map
Stable entry URLs from the Agents hub into Start, Tutorials, Guides, Reference, Enterprise, and Contribute.
- Agents hubThis page is the Agents and Cursor navigation hub.
- StartInstall LCDL, adoption paths, and troubleshooting.
- Tutorials (101 / 201 / 301)Progressive labs from introductory through advanced topics.
- GuidesPlaywright discovery, task packs, and graph workflows.
- ReferenceClient API, ContractSpec, schemas, and task catalog.
- EnterpriseThreat model, operations, and compatibility expectations.
- ContributeDocs authoring, release checklist, and website publishing.
Agents hub
|
+-- Start
+-- Tutorials (101 / 201 / 301)
+-- Guides
+-- Reference
+-- Enterprise
+-- ContributeRelated: Tutorial 301 (enterprise + MCP operations), Client API.