feat(codegen): add docs generation for ORM, React Query, and multi-target support

[pyramation]

feat(codegen): extend docs generation to ORM and React Query generators

Summary

Extends the docs generation system (previously CLI-only from PR #725) to also cover ORM and React Query generators. Adds multi-target orchestration (generateMulti()) to core so the CLI entry point delegates looping rather than owning it.

New files:

• docs-utils.ts — shared utilities extracted from the CLI docs generator (resolveDocsConfig, fieldTypeToTs, McpTool type)
• orm/docs-generator.ts — README, AGENTS.md, MCP tools, and skills for the ORM layer (~525 lines)
• hooks-docs-generator.ts — README, AGENTS.md, MCP tools, and skills for React Query hooks (~595 lines)
• target-docs-generator.ts — per-target root README, combined mcp.json, root-root README for multi-target configs

Modified files:

• types/config.ts — docs: DocsConfig | boolean moved to top-level GraphQLSDKConfigTarget (removed from CliConfig)
• core/generate.ts — wires docs generation for all enabled generators after code generation; adds generateMulti() for multi-target configs
• cli/index.ts (entry point) — uses generateMulti() instead of its own loop
• cli/docs-generator.ts — refactored to import shared utils from docs-utils.ts
• cli/index.ts (codegen) — removed docs generation from generateCli(); docs are now handled centrally in generate()

3-level README hierarchy:

1. Root-root README (multi-target only) — table-of-contents routing to each target
2. Per-target README — ties together ORM/hooks/CLI docs for one API
3. Per-generator README/AGENTS.md/skills — detailed docs per generator subdirectory

Review & Testing Checklist for Human

• generateMulti() file writing approach — the root-root README is written via a dynamic import('./output') and writeGeneratedFiles() call to . (outside the normal filesToWrite array flow). Verify this doesn't cause issues with pruning, path resolution, or concurrent writes in real multi-target configs.
• Docs generators are all string-based markdown — unlike the code generators which use Babel AST, these produce markdown via string concatenation. The previous conversation had a "no string concatenation" directive for code generation. Verify this is acceptable for non-code (markdown/JSON) output, or if you'd prefer a different approach.
• Lockfile changes — pnpm-lock.yaml has thousands of lines changed including @types/node version shifts (22 → 20) and ts-node entry removals. Verify this was intentional or if the lockfile was regenerated in a different environment. May want to regenerate with pnpm install in a clean state.
• Test a multi-target config end-to-end — create a config with 3 targets (e.g., auth, members, app) each with different generator combinations, run codegen, verify:
  • Each target's output directory has correct docs
  • Root-root README is generated at project root with correct links
  • Combined mcp.json at each target root includes tools from all enabled generators
  • No file path issues or missing docs

Notes

• All 262 tests pass, 31 new snapshots added
• Defaults: readme + agents always on, mcp + skills opt-in
• docs: true enables all 4 formats
• docs: false disables all docs

Link to Devin run: https://app.devin.ai/sessions/e9c668834d394cdc8ebcb371b6ebf430
Requested by: @pyramation

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring