Skip to content

docs(core): refine IA — Quickstart per agent + question-titled concepts#2311

Merged
goldmedal merged 1 commit into
mainfrom
feat/refine-docs
May 20, 2026
Merged

docs(core): refine IA — Quickstart per agent + question-titled concepts#2311
goldmedal merged 1 commit into
mainfrom
feat/refine-docs

Conversation

@PaulChen79
Copy link
Copy Markdown
Contributor

@PaulChen79 PaulChen79 commented May 20, 2026

Summary

Major IA restructure of docs/core/. The companion doc-site PR is Canner/docs.getwren.ai#315 — after both land, the sync-docs workflow will mirror this content over.

What changed

Get Started

  • installation.md — kept as the flat skill-bundle install (npx skills add ... + onboarding-skill prompt)
  • quickstart-with-agent/ — new folder with 53 per-agent install pages (Claude Code, Cursor, Codex, Continue, Cline, Cursor, Goose, Windsurf, OpenCode, Roo, …). Each page has prerequisite, the canonical npx skills add Canner/WrenAI --skill '*' --agent <flag> command, an agent-specific session-start instruction (CLI / IDE / cloud), and the /wren-onboarding slash-command prompt
  • quickstart.md — link updates only

Concepts (6 question-titled pages)

Page Status
what_is_context.md kept, link fixes only
agent_learning.md new — scaffold/grill/auto-pilot + skills + iteration loop
memory_system.md content unchanged, link fixes
what_is_mdl.md content unchanged, link fixes
correctness.md new — 6 correctness pillars + cubes + dry-plan
stack_position.md new — layering on warehouse / transformation / existing semantic layer

Deleted: concepts/skills.md, concepts/architecture.md (both moved to reference/).

Guides (5 outcome recipe cards)

Card Status
connect.md kept, light tweak
model.md new — agent-driven MDL generation outcome
refine.md new — memory + instructions + iteration loop (notes wren-enrich-context is in active development)
cubes.md new — pre-aggregation primitive
manage_project.md new — project lifecycle + profile workflow + multi-env

Deleted: guides/modeling/ (6 files — field-level spec moved to reference/mdl.md), guides/memory.md, guides/profiles.md, guides/use-with-agents/claude-code.md.

Reference (5 atomic pages)

Page Status
cli.md link fix only
skills.md link fix only
mdl.md new — consolidates project + models + relationships + views + cubes + instructions field specs
architecture.md moved from concepts/ — renamed WrenEngine section to Wren Python SDK (per-review feedback); mermaid simplified to avoid layout overflow
operational.md new — every path, env var, discovery rule on one page (Senior DE bookmark page)

Overview (introduction.mdx)

Light edits: deep-link each section into the matching Concept page; merged "How it fits your stack" into "How you use it"; "Wren AI Core" → "Wren AI".

Test plan

  • CI passes (no docs lint job today but worth eyeballing)
  • After merge, sync-docs workflow opens a PR on docs.getwren.ai
  • After both merge + deploy, walk through the new sidebar on docs.getwren.ai

🤖 Generated with Claude Code

Summary by CodeRabbit

  • Documentation
    • Added 48+ agent-specific quickstart guides (AdaL, Claude Code, GitHub Copilot, Continue, Cursor, etc.) with consistent installation and onboarding instructions.
    • Introduced comprehensive conceptual guides on agent learning, system correctness, and architectural positioning.
    • Added new reference documentation for MDL schema, operational setup, and system architecture.
    • Reorganized and expanded guides for project management, model creation, and cube queries.

Review Change Stack

Major restructure aligned with docs.getwren.ai PR #315:

- Get Started: add quickstart-with-agent/ with 53 supported agents
  (Claude Code, Cursor, Codex, Continue, Cline, Roo, Windsurf, ...);
  each page has prerequisites, install command, start-session instruction
  (CLI / IDE / cloud variant), and the canonical onboarding-skill prompt
- Concepts: 6 question-titled pages — what_is_context, agent_learning (new),
  memory_system, what_is_mdl, correctness (new), stack_position (new)
- Guides: 5 outcome recipe cards — connect, model (new), refine (new),
  cubes (new), manage_project (new); deleted modeling/ subfolder
  (field-level detail moved to reference/mdl.md), memory.md, profiles.md,
  use-with-agents/
- Reference: add mdl.md (consolidates project + models + relations + views
  + cubes + instructions), architecture.md (moved from concepts/, renamed
  WrenEngine section to Wren Python SDK), operational.md (paths + env
  vars + discovery rules)
- Overview: tease six concept pages with deep-links; merge "How it fits
  your stack" into "How you use it"

The doc site sidebar + redirects to match are in Canner/docs.getwren.ai
PR #315. After this merges, sync-docs workflow will propagate the
docs/ changes there.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@github-actions github-actions Bot added the documentation Improvements or additions to documentation label May 20, 2026
@coderabbitai
Copy link
Copy Markdown
Contributor

coderabbitai Bot commented May 20, 2026

Caution

Review failed

The pull request is closed.

ℹ️ Recent review info
⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 53c7b871-a172-4fb0-ae2b-ed7376fd30d5

📥 Commits

Reviewing files that changed from the base of the PR and between eaedc8d and 495b9e3.

📒 Files selected for processing (83)
  • docs/core/concepts/agent_learning.md
  • docs/core/concepts/architecture.md
  • docs/core/concepts/correctness.md
  • docs/core/concepts/memory_system.md
  • docs/core/concepts/skills.md
  • docs/core/concepts/stack_position.md
  • docs/core/concepts/what_is_context.md
  • docs/core/concepts/what_is_mdl.md
  • docs/core/get_started/installation.md
  • docs/core/get_started/quickstart-with-agent/adal.md
  • docs/core/get_started/quickstart-with-agent/aider-desk.md
  • docs/core/get_started/quickstart-with-agent/amp.md
  • docs/core/get_started/quickstart-with-agent/antigravity.md
  • docs/core/get_started/quickstart-with-agent/augment.md
  • docs/core/get_started/quickstart-with-agent/bob.md
  • docs/core/get_started/quickstart-with-agent/claude-code.md
  • docs/core/get_started/quickstart-with-agent/cline.md
  • docs/core/get_started/quickstart-with-agent/codearts-agent.md
  • docs/core/get_started/quickstart-with-agent/codebuddy.md
  • docs/core/get_started/quickstart-with-agent/codemaker.md
  • docs/core/get_started/quickstart-with-agent/codestudio.md
  • docs/core/get_started/quickstart-with-agent/codex.md
  • docs/core/get_started/quickstart-with-agent/command-code.md
  • docs/core/get_started/quickstart-with-agent/continue.md
  • docs/core/get_started/quickstart-with-agent/cortex.md
  • docs/core/get_started/quickstart-with-agent/crush.md
  • docs/core/get_started/quickstart-with-agent/cursor.md
  • docs/core/get_started/quickstart-with-agent/deepagents.md
  • docs/core/get_started/quickstart-with-agent/devin.md
  • docs/core/get_started/quickstart-with-agent/dexto.md
  • docs/core/get_started/quickstart-with-agent/droid.md
  • docs/core/get_started/quickstart-with-agent/firebender.md
  • docs/core/get_started/quickstart-with-agent/forgecode.md
  • docs/core/get_started/quickstart-with-agent/gemini-cli.md
  • docs/core/get_started/quickstart-with-agent/github-copilot.md
  • docs/core/get_started/quickstart-with-agent/goose.md
  • docs/core/get_started/quickstart-with-agent/hermes-agent.md
  • docs/core/get_started/quickstart-with-agent/iflow-cli.md
  • docs/core/get_started/quickstart-with-agent/junie.md
  • docs/core/get_started/quickstart-with-agent/kilo.md
  • docs/core/get_started/quickstart-with-agent/kimi-cli.md
  • docs/core/get_started/quickstart-with-agent/kiro-cli.md
  • docs/core/get_started/quickstart-with-agent/kode.md
  • docs/core/get_started/quickstart-with-agent/mcpjam.md
  • docs/core/get_started/quickstart-with-agent/mistral-vibe.md
  • docs/core/get_started/quickstart-with-agent/mux.md
  • docs/core/get_started/quickstart-with-agent/neovate.md
  • docs/core/get_started/quickstart-with-agent/openclaw.md
  • docs/core/get_started/quickstart-with-agent/opencode.md
  • docs/core/get_started/quickstart-with-agent/openhands.md
  • docs/core/get_started/quickstart-with-agent/pi.md
  • docs/core/get_started/quickstart-with-agent/pochi.md
  • docs/core/get_started/quickstart-with-agent/qoder.md
  • docs/core/get_started/quickstart-with-agent/qwen-code.md
  • docs/core/get_started/quickstart-with-agent/roo.md
  • docs/core/get_started/quickstart-with-agent/rovodev.md
  • docs/core/get_started/quickstart-with-agent/tabnine-cli.md
  • docs/core/get_started/quickstart-with-agent/trae-cn.md
  • docs/core/get_started/quickstart-with-agent/trae.md
  • docs/core/get_started/quickstart-with-agent/warp.md
  • docs/core/get_started/quickstart-with-agent/windsurf.md
  • docs/core/get_started/quickstart-with-agent/zencoder.md
  • docs/core/get_started/quickstart.md
  • docs/core/guides/connect.md
  • docs/core/guides/cubes.md
  • docs/core/guides/manage_project.md
  • docs/core/guides/memory.md
  • docs/core/guides/model.md
  • docs/core/guides/modeling/cube.md
  • docs/core/guides/modeling/model.md
  • docs/core/guides/modeling/overview.md
  • docs/core/guides/modeling/relation.md
  • docs/core/guides/modeling/view.md
  • docs/core/guides/modeling/wren_project.md
  • docs/core/guides/profiles.md
  • docs/core/guides/refine.md
  • docs/core/guides/use-with-agents/claude-code.md
  • docs/core/introduction.mdx
  • docs/core/reference/architecture.md
  • docs/core/reference/cli.md
  • docs/core/reference/mdl.md
  • docs/core/reference/operational.md
  • docs/core/sdk/wasm.md

Walkthrough

This PR is a comprehensive documentation restructure for the Wren AI project, rewriting conceptual narrative around the "context layer" model, adding 50+ agent integration guides, reorganizing workflow-focused guides (replacing the old modeling structure), and establishing new reference documentation for architecture and operational configuration. No code changes.

Changes

Core Concept & Introduction Rewrite

Layer / File(s) Summary
What is context and MDL: semantic layer foundations
docs/core/concepts/what_is_context.md, docs/core/concepts/what_is_mdl.md
Context redefined as full machine-readable business understanding across five layers (structural, semantic, business, operational, behavioral). MDL repositioned as a semantic contract with emphasis on portability, project structure, and execution mapping to the Rust semantic engine.
Correctness & memory: system pillars and behavioral layer
docs/core/concepts/correctness.md, docs/core/concepts/memory_system.md
Correctness framed as six interdependent pillars (schema linking, value profiling, ambiguity detection, generation trace, retry/repair, eval). Memory reframed as behavioral/retrieval layer with two-collection index (schema_items, query_history) and explicit grounding loop (recall → fetch → execute → store).
Agent learning loop & stack positioning
docs/core/concepts/agent_learning.md, docs/core/concepts/stack_position.md
Introduces agent learning persistence across MDL, instructions, memory, and queries.yml. Clarifies Wren AI's role as context layer between data infrastructure and agents.
Introduction & product narrative
docs/core/introduction.mdx
Restructures product positioning from "what it does" to "why it exists / what it provides." Adds five-layer context table, condensed correctness explanation, "scaffold fast, enrich deep" workflow, and updated navigation.

Agent Integration Quickstarts

Layer / File(s) Summary
Installation guide & quickstart base updates
docs/core/get_started/installation.md, docs/core/get_started/quickstart.md
Updates to reference /wren-onboarding skill explicitly and fixes internal link targets to new guide locations (MDL, cubes).
Agent quickstarts: 50+ integrations
docs/core/get_started/quickstart-with-agent/*.md (AdaL, AiderDesk, Amp, Antigravity, Augment, Bob, Claude Code, Cline, CodeArts, CodeBuddy, Codemaker, CodeStudio, Codex, Command Code, Continue, Cortex, Crush, Cursor, Deep Agents, Devin, Dexto, Droid, Firebender, ForgeCode, Gemini CLI, GitHub Copilot, Goose, Hermes, iFlow, Junie, Kilo, Kimi, Kiro, Kode, MCPJam, Mistral Vibe, Mux, Neovate, OpenClaw, OpenCode, OpenHands, Pi, Pochi, Qoder, Qwen Code, Roo, Rovo Dev, Tabnine, Trae CN, Trae, Warp, Windsurf, Zencoder)
Adds comprehensive quickstart pages for 50+ agents, each following a consistent template: prerequisites, skill installation via npx skills add, and /wren-onboarding onboarding workflow.

Guides Restructuring: From Modeling to Workflows

Layer / File(s) Summary
Remove old modeling guide structure
docs/core/guides/modeling/*, docs/core/guides/profiles.md, docs/core/guides/memory.md, docs/core/guides/use-with-agents/claude-code.md
Removes deprecated modeling guides (overview, model, relation, view, cube, wren_project), legacy profiles guide, and old memory guide and Claude Code integration.
New workflow-centric guides
docs/core/guides/model.md, docs/core/guides/manage_project.md, docs/core/guides/refine.md, docs/core/guides/cubes.md, docs/core/guides/connect.md
Adds step-by-step MDL modeling guide, comprehensive project management (lifecycle, profiles, multi-environment, secrets), iterative quality refinement loop, and cube semantics. Updates navigation in connect.md.

Reference Documentation & Link Updates

Layer / File(s) Summary
Architecture & operational reference
docs/core/reference/architecture.md, docs/core/reference/operational.md
New architecture.md describes end-to-end system design with correctness primitives, four-layer model, component details (skills, CLI, SDK, planning, semantic engine, memory), and data-flow diagrams. New operational.md covers filesystem layout, environment variables, project/connection discovery, and secret handling.
MDL schema reference documentation
docs/core/reference/mdl.md
Comprehensive YAML schema reference documenting wren_project.yml, models, relationships, views, cubes, instructions, and queries.yml with field tables, examples, validation rules, and transformation details.
CLI and SDK link updates
docs/core/reference/cli.md, docs/core/sdk/wasm.md
Updates cube-related documentation links from removed docs/core/guides/modeling/cube.md to new docs/core/guides/cubes.md.

🎯 2 (Simple) | ⏱️ ~15 minutes

Possibly Related PRs

  • Canner/WrenAI#2241: Part of the same docs/core structure rebrand and reorganization.
  • Canner/WrenAI#2285: Shares the same "correctness-as-a-system" and context-layer narrative restructuring.
  • Canner/WrenAI#2263: Overlaps in memory documentation changes and context layer positioning.

Suggested Labels

documentation, skills

Suggested Reviewers

  • goldmedal
  • cyyeh
  • chilijung

🐰 Hopping through concepts and quickstarts with glee,
Fifty agents now ready to help users be free,
From context to correctness, the story runs deep,
A well-structured docs garden for all to keep! 🌿✨

✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests
  • Commit unit tests in branch feat/refine-docs

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands and usage tips.

@goldmedal goldmedal merged commit dbca75a into main May 20, 2026
2 of 3 checks passed
@goldmedal goldmedal deleted the feat/refine-docs branch May 20, 2026 11:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

documentation Improvements or additions to documentation

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants