Phase 4: Ask Petal SSE chat

Conversational follow-up on a suggestion, streamed token-by-token.

Backend (interface-only; handlers never touch a concrete LLM client):
- internal/llm/chat.go: StreamAskPetal with conversational sampling
  (max_tokens 512, temp 0.7, rep 1.15, top_p 0.92, stop "\n\n\n"),
  reusing AskPetalSystemPrompt + TrimHistory.
- internal/suggestions/chat.go: POST /api/suggestions/:id/chat. One
  user-scoped join loads the suggestion + parent content_text;
  surroundingParagraph extracts the \n\n-bounded paragraph at from_pos
  (whole-doc fallback when unlocated) and injects it server-side.
  Streams event: token / event: done SSE frames with JSON-encoded data
  so token newlines can't break framing; real http.Flusher per chunk.
  LLM-unreachable -> 502 before SSE headers; unknown suggestion -> 404.

Frontend:
- streamSuggestionChat: fetch + ReadableStream SSE parser (not
  EventSource, needs POST), abortable.
- AskPetal.tsx: whole conversation in component state (no persistence,
  cleared on close), Petal's first bubble pre-seeded with the
  explanation, rose/lavender bubbles, CJK font stack on the bubbles
  only (Note #17), streaming caret.
- SuggestionCard "Ask Petal" pill pins the card open while chatting
  (hover-close suppressed, click-away closes) and widens it to 340px.

Tests: chat_test.go covers streamed-text concat + done event,
server-side context injection on the system message, sampling params,
404, and surroundingParagraph. go build/vet/test clean, tsc clean,
vite build OK. Live SSE smoke-tested against a fake streaming vLLM:
tokens flushed individually through the chi middleware stack, done
terminator, 502 on LLM-down, 404 on unknown suggestion.

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

web/src/api/client.ts

@@ -78,3 +78,64 @@ export const api = {
   dismissSuggestion: (id: string) =>
     req<void>(`/suggestions/${id}/dismiss`, { method: 'POST' }),
 }
+
+// One turn in an Ask Petal conversation. History lives only in the component —
+// the server is stateless and re-injects the suggestion context every request.
+export interface ChatMessage {
+  role: 'user' | 'assistant'
+  content: string
+}
+
+// streamSuggestionChat POSTs the conversation to the SSE chat endpoint and
+// invokes onToken for each text chunk as it arrives. It uses fetch + a
+// ReadableStream reader (not EventSource, which can't POST) and resolves when
+// the stream ends. Abort via the optional signal to cancel mid-response.
+export async function streamSuggestionChat(
+  suggestionId: string,
+  messages: ChatMessage[],
+  onToken: (text: string) => void,
+  signal?: AbortSignal,
+): Promise<void> {
+  const res = await fetch(`/api/suggestions/${suggestionId}/chat`, {
+    method: 'POST',
+    headers: { 'Content-Type': 'application/json' },
+    body: JSON.stringify({ messages }),
+    signal,
+  })
+  if (!res.ok || !res.body) {
+    const detail = await res.text().catch(() => '')
+    throw new Error(`${res.status} ${res.statusText}${detail ? `: ${detail}` : ''}`)
+  }
+
+  const reader = res.body.getReader()
+  const decoder = new TextDecoder()
+  let buf = ''
+  for (;;) {
+    const { done, value } = await reader.read()
+    if (done) break
+    buf += decoder.decode(value, { stream: true })
+    // SSE events are separated by a blank line. Process every complete one and
+    // keep the trailing partial in the buffer.
+    let sep: number
+    while ((sep = buf.indexOf('\n\n')) >= 0) {
+      const event = parseSSE(buf.slice(0, sep))
+      buf = buf.slice(sep + 2)
+      if (event.name === 'done') return
+      if (event.name === 'token' && event.data) {
+        const text = (JSON.parse(event.data) as { text: string }).text
+        if (text) onToken(text)
+      }
+    }
+  }
+}
+
+// parseSSE pulls the event name and data payload out of one raw SSE frame.
+function parseSSE(frame: string): { name: string; data: string } {
+  let name = 'message'
+  const dataLines: string[] = []
+  for (const line of frame.split('\n')) {
+    if (line.startsWith('event:')) name = line.slice(6).trim()
+    else if (line.startsWith('data:')) dataLines.push(line.slice(5).trim())
+  }
+  return { name, data: dataLines.join('\n') }
+}