feat: per-job token-usage capture via PostToolUse hook

update_job_status accepts optionele model_id + 4 token-velden conform het
runbook-contract (mcp-integration.md:42). De waarden komen niet van de agent
zelf maar van scripts/persist-job-usage.ts, een PostToolUse-hook die het
lokale Claude Code transcript (~/.claude/projects/.../*.jsonl) leest en de
usage tussen de laatste wait_for_job en update_job_status optelt.

Geen Anthropic API-key nodig — alle data staat al lokaal op disk omdat
Claude Code per assistant-message het API usage-blok logt
(input_tokens, output_tokens, cache_creation_input_tokens,
cache_read_input_tokens + message.model).

Robustness:
- Subagent (isSidechain: true) lines worden geskipt om double-counting
  te voorkomen tegen subagents/-subdirectory transcripts.
- Lines worden gededupliceerd op uuid (branching/resumption).
- model_id wordt genormaliseerd: claude-opus-4-7[1m] -> claude-opus-4-7-1m
  zodat de [1m]-variant op een aparte model_prices-rij kan matchen.
- Hook is non-blocking: elke fout logt een warning en exit 0.

Hook-config in .claude/settings.json met SCRUM4ME_MCP_DIR-fallback zodat
de agent vanuit een product-worktree (andere cwd) ook werkt mits de user
de hook in ~/.claude/settings.json kopieert.

16 nieuwe vitest-cases voor parseTranscript, computeUsageFromTranscript,
normalizeModelId en persistJobUsage.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src/tools/update-job-status.ts

@@ -21,6 +21,11 @@ const inputSchema = z.object({
   branch: z.string().min(1).optional(),
   summary: z.string().max(1_000).optional(),
   error: z.string().max(2_000).optional(),
+  model_id: z.string().min(1).max(200).optional(),
+  input_tokens: z.number().int().nonnegative().optional(),
+  output_tokens: z.number().int().nonnegative().optional(),
+  cache_read_tokens: z.number().int().nonnegative().optional(),
+  cache_write_tokens: z.number().int().nonnegative().optional(),
 })
 
 export async function cleanupWorktreeForTerminalStatus(
@@ -266,10 +271,24 @@ export function registerUpdateJobStatusTool(server: McpServer) {
         'PARTIAL/DIVERGENT but requires a non-empty summary (≥20 chars) explaining the drift; ANY ' +
         'accepts everything. ' +
         'Automatically emits an SSE event so the Scrum4Me UI updates in real time. ' +
+        'Optionally accepts token-usage fields (model_id + input/output/cache_read/cache_write tokens) ' +
+        'for cost tracking — typically populated by a PostToolUse hook from the local Claude Code transcript, ' +
+        'not by the agent itself. ' +
         'Response includes next_action: when wait_for_job_again, immediately call wait_for_job again. When queue_empty, the agent batch is done.',
       inputSchema,
     },
-    async ({ job_id, status, branch, summary, error }) =>
+    async ({
+      job_id,
+      status,
+      branch,
+      summary,
+      error,
+      model_id,
+      input_tokens,
+      output_tokens,
+      cache_read_tokens,
+      cache_write_tokens,
+    }) =>
       withToolErrors(async () => {
         const auth = await requireWriteAccess()
         const { tokenId, userId } = auth
@@ -371,6 +390,11 @@ export function registerUpdateJobStatusTool(server: McpServer) {
             ...(summary !== undefined ? { summary } : {}),
             ...(errorToWrite !== undefined ? { error: errorToWrite } : {}),
             ...(prUrl !== null ? { pr_url: prUrl } : {}),
+            ...(model_id !== undefined ? { model_id } : {}),
+            ...(input_tokens !== undefined ? { input_tokens } : {}),
+            ...(output_tokens !== undefined ? { output_tokens } : {}),
+            ...(cache_read_tokens !== undefined ? { cache_read_tokens } : {}),
+            ...(cache_write_tokens !== undefined ? { cache_write_tokens } : {}),
           },
           select: {
             id: true,