Phase 3: LLM grammar checkpoint

Backend (internal/llm): backend-agnostic LLMClient interface + factory
with vLLM (OpenAI-compat) and Ollama (native) clients, each Complete +
Stream. prompts.go holds the checkpoint and Ask Petal templates;
checkpoint.go salvages JSON from model output (brace-matched), enforces a
per-doc 30s RateLimiter, and truncates the doc to a latency cap.

internal/suggestions: POST /api/docs/:id/check runs a checkpoint and
replaces the doc's pending suggestions in one tx (accepted/rejected kept
as history); GET /api/docs/:id/suggestions lists pending;
POST /api/suggestions/:id/{accept,dismiss} resolves one. Throttled checks
return the current set rather than erroring.

Frontend: useCheckpoint (4s debounce, loads existing on open, stale-guard
tokens); SuggestionHighlight renders ProseMirror decorations re-anchored
by the `original` string on every doc change (not stored marks), with
precise textblock-offset→PM-position mapping; SuggestionCard shows the
type tag + diff + explanation and applies the replacement in-editor on
accept; breathing rose checkpoint dot in the StatusBar; fade-float +
breathe animations.

Tests: llm parse/rate-limit/truncate; suggestions full flow + rate-limit
over httptest with a stub client. Smoke-tested end-to-end against a fake
vLLM endpoint (anchoring verified) and the LLM-unreachable 502 path.

Claude-Session: https://claude.ai/code/session_016Yr6jELuRc7hyzYLccQKZd

web/src/api/client.ts

@@ -28,6 +28,25 @@ export interface DocUpdate {
   word_count?: number
 }
 
+export type SuggestionType = 'grammar' | 'phrasing' | 'idiom' | 'clarity' | 'voice'
+
+// A single LLM-proposed edit. `original` is the source of truth for placement —
+// the editor re-anchors by matching this string in the live document (spec
+// Note #6); from_pos/to_pos are server-side advisory only. `replacement` is
+// empty for voice flags (awareness-only, no correction to apply).
+export interface Suggestion {
+  id: string
+  doc_id: string
+  from_pos: number
+  to_pos: number
+  original: string
+  replacement: string
+  explanation: string
+  type: SuggestionType
+  status: 'pending' | 'accepted' | 'rejected'
+  created_at: string
+}
+
 async function req<T>(path: string, init?: RequestInit): Promise<T> {
   const res = await fetch(`/api${path}`, {
     headers: { 'Content-Type': 'application/json' },
@@ -48,4 +67,14 @@ export const api = {
   updateDoc: (id: string, body: DocUpdate) =>
     req<Document>(`/docs/${id}`, { method: 'PUT', body: JSON.stringify(body) }),
   deleteDoc: (id: string) => req<void>(`/docs/${id}`, { method: 'DELETE' }),
+
+  // Grammar checkpoint: runs an LLM pass and returns the fresh pending set.
+  // Rate-limited per document server-side (returns the existing set if too soon).
+  checkDoc: (id: string) => req<Suggestion[]>(`/docs/${id}/check`, { method: 'POST' }),
+  // Pending suggestions for a doc, loaded when the editor opens it.
+  listSuggestions: (id: string) => req<Suggestion[]>(`/docs/${id}/suggestions`),
+  acceptSuggestion: (id: string) =>
+    req<void>(`/suggestions/${id}/accept`, { method: 'POST' }),
+  dismissSuggestion: (id: string) =>
+    req<void>(`/suggestions/${id}/dismiss`, { method: 'POST' }),
 }