@fastagent-sh/voicenote 0.22.0 → 0.22.1

package/README.md

@@ -215,7 +215,7 @@ Changes take effect on the next `vn run`. `~`, `$HOME`, and `${HOME}` are accept
 6. The summary model (default: pi codex via ChatGPT Plus) reads the raw transcript directly, performing necessary cleanup, speaker restoration, and reconstruction of views/debates/consensus inside the notes-generation stage; if the summary fails, the next `vn run` / `vn run --latest` reuses the saved transcript and retries only the notes generation — no `vn forget` needed
 7. Write notes / metadata; the system makes no archiving decisions — files stay in the configured workspace
 
-A failing recording is retried on later runs, but at most **3 times** (whether it fails in transcription or in summarisation, and a run killed mid-job counts too). After that it is marked `Gave up` and left alone, so one broken file can't burn ASR/LLM budget on every scheduler tick — `vn forget <name>` drops the record and re-queues it. Re-queuing is not the same as re-transcribing: if the transcript is already on disk it is reused, so `vn forget` never re-pays for ASR. (`vn forget` takes the run lock, so it refuses while a run is in progress — wait for that run to finish and repeat.)
+A failing recording is retried on later runs, but at most **3 times** (whether it fails in transcription or in summarisation, and a run killed mid-job counts too). After that it is marked `Gave up` and left alone, so one broken file can't burn ASR/LLM budget on every scheduler tick. Use **Retry** on its GUI row to reset the budget, preserve saved outputs, and run it again; `vn forget <name>` is the CLI escape hatch that drops the record and re-queues it. Either path reuses a saved transcript instead of paying for ASR again. Both take the run lock, so retry after the active run finishes if the state file is busy.
 
 Records whose source file is no longer on the recorder are forgotten on the next scan (and the removal is logged), *unless* they already produced notes or a transcript — that history is kept. This is why swapping recorders, or deleting files from the device, no longer leaves permanent "failed" rows behind.
 
@@ -227,7 +227,7 @@ Records whose source file is no longer on the recorder are forgotten on the next
 - Original audio: `${VOICENOTE_WORKSPACE}/_audio/YYYY-MM/`
 - Full transcripts: `${VOICENOTE_WORKSPACE}/_transcripts/YYYY-MM/`
 - Metadata: `${VOICENOTE_WORKSPACE}/_metadata/YYYY-MM/`
-- State: `${VOICENOTE_WORKSPACE}/_state/jobs.json` — one record per recording, holding its `state` — where it is in its lifecycle (`queued`, `running`, `done`, `filtered`, `error`, or `gave_up` once retries are spent) — plus a `code` saying why (`summary_failed`, `transcribe_failed`, `interrupted`, `too_small`, …), its attempt count and its output paths. `vn run` is the only writer; `vn jobs` and the GUI dashboard are pure reads of it, so what you see is what will run. A pre-0.18 `processed.json` is converted automatically on the first run and kept as `processed.json.v1.bak`.
+- State: `${VOICENOTE_WORKSPACE}/_state/jobs.json` — one record per recording, holding its `state` — where it is in its lifecycle (`queued`, `running`, `done`, `filtered`, `error`, or `gave_up` once retries are spent) — plus a `code` saying why (`summary_failed`, `transcribe_failed`, `interrupted`, `too_small`, …), its attempt count and its output paths. `vn run` writes lifecycle updates; only explicit retry/forget actions mutate it otherwise. `vn jobs` and passive GUI refreshes are pure reads, so what you see is what will run. A pre-0.18 `processed.json` is converted automatically on the first run and kept as `processed.json.v1.bak`.
 - Index: `${VOICENOTE_WORKSPACE}/_index/notes.jsonl`
 
 ## Automation
@@ -282,10 +282,10 @@ The workflow lives at `.github/workflows/release.yml`: CI explicitly runs typech
 
 A self-contained macOS `.app` (Tauri v2) for **non-terminal users**: the target machine needs no pre-installed bun / pi / ffprobe / global `vn`.
 
-**Positioning**: the GUI is only a "status dashboard + quick access to output" — it does **not** drive processing. The full pipeline runs autonomously every 60s via the background LaunchAgent using the bundled CLI (it keeps running with the GUI closed).
+**Positioning**: the GUI is a status dashboard with quick access to output and manual Sync/Retry controls. The full pipeline still runs autonomously every 60s via the background LaunchAgent using the bundled CLI (it keeps running with the GUI closed).
 
 - First run: settings (identity / Volcano keys / proxy). The notes model comes from pi; ChatGPT users can sign in from the Status panel (`vn login`'s browser-callback flow).
-- After that: the main view shows agent activity + recent notes (open note / open folder)
+- After that: the main view shows agent activity and recent notes, opens outputs, and can retry failed recordings.
 
 ### What's bundled
 

package/README.zh-CN.md

@@ -210,7 +210,7 @@ vn uninstall-launch-agent
 6. summary 模型(由 pi 自身配置决定)直接看原始 transcript,在纪要生成阶段内部完成必要清理、说话人还原、观点/争论/共识形成过程还原;如果 summary 失败,下一次 `vn run` / `vn run --latest` 会复用已保存 transcript,直接重试纪要生成,不需要 `vn forget`
 7. 写出 notes / metadata；系统不做任何归档决定，文件留在配置的 workspace 中
 
-失败的录音会在后续运行中重试，但**最多 3 次**（转写失败、纪要失败、以及被中途 kill 的运行都算）。超过后标记为 `Gave up` 并不再自动重试，避免一个坏文件每个调度周期都烧一次 ASR/LLM 额度 —— `vn forget <name>` 会删掉该记录并重新入队。重新入队不等于重新转写：磁盘上已有 transcript 时会直接复用，所以 `vn forget` 不会让你再付一次 ASR。（`vn forget` 需要 run lock，因此在某次 run 进行中时会拒绝执行 —— 等该次 run 结束后重试即可。）
+失败的录音会在后续运行中重试，但**最多 3 次**（转写失败、纪要失败、以及被中途 kill 的运行都算）。超过后标记为 `Gave up` 并不再自动重试，避免一个坏文件每个调度周期都烧一次 ASR/LLM 额度。点击 GUI 记录上的**重试**会重置次数、保留已有产物并立即再跑；CLI 也可以用 `vn forget <name>` 删除记录后重新入队。两种方式都会复用磁盘上已有的 transcript，不会重复支付 ASR 费用。它们都需要 run lock；如果当前正在处理，请等本次 run 结束后再重试。
 
 源文件已不在录音笔上的记录，会在下一次扫描时被遗忘（并记入日志），**已经产出纪要或 transcript 的除外** —— 那部分历史会保留。所以换录音笔、或从设备上删文件，不再会留下永久的 “失败” 条目。
 
@@ -222,7 +222,7 @@ vn uninstall-launch-agent
 - 原始音频:`${VOICENOTE_WORKSPACE}/_audio/YYYY-MM/`
 - 完整转写:`${VOICENOTE_WORKSPACE}/_transcripts/YYYY-MM/`
 - metadata:`${VOICENOTE_WORKSPACE}/_metadata/YYYY-MM/`
-- 状态：`${VOICENOTE_WORKSPACE}/_state/jobs.json` —— 每条录音一条记录，包含 `state`（生命周期位置：`queued`、`running`、`done`、`filtered`、`error`，以及重试耗尽后的 `gave_up`）、`code`（原因：`summary_failed`、`transcribe_failed`、`interrupted`、`too_small` 等）、重试次数和产物路径。`vn run` 是唯一的写入方，`vn jobs` 和 GUI 面板都只是它的纯读取 —— 你看到的队列就是会跑的队列。0.18 之前的 `processed.json` 会在首次运行时自动转换，旧文件保留为 `processed.json.v1.bak`。
+- 状态：`${VOICENOTE_WORKSPACE}/_state/jobs.json` —— 每条录音一条记录，包含 `state`（生命周期位置：`queued`、`running`、`done`、`filtered`、`error`，以及重试耗尽后的 `gave_up`）、`code`（原因：`summary_failed`、`transcribe_failed`、`interrupted`、`too_small` 等）、重试次数和产物路径。`vn run` 写入正常生命周期变化，除此之外只有显式重试或 forget 操作会修改它；`vn jobs` 和 GUI 的被动刷新都是纯读取，因此看到的队列就是实际会跑的队列。0.18 之前的 `processed.json` 会在首次运行时自动转换，旧文件保留为 `processed.json.v1.bak`。
 - 索引:`${VOICENOTE_WORKSPACE}/_index/notes.jsonl`
 
 ## 自动化
@@ -277,10 +277,10 @@ workflow 位于 `.github/workflows/release.yml`:CI 显式跑 typecheck、测试
 
 面向**非终端用户**:一个自包含的 macOS `.app`(Tauri v2),目标机器无需预装 bun / pi / ffprobe / 全局 `vn`。
 
-**定位**:GUI 只是「工作状态 dashboard + 产出快捷入口」,**不驱动处理**。真正的全流程由后台 LaunchAgent 用包内 CLI 每 60s 自主运行(关掉 GUI 也跑)。
+**定位**:GUI 是工作状态 dashboard，提供产出快捷入口以及手动同步/重试。全流程仍由后台 LaunchAgent 用包内 CLI 每 60s 自主运行(关掉 GUI 也跑)。
 
 - 首次:配置向导(身份 / Volcano keys / 代理)→ ChatGPT 登录(设备无终端,走 `vn login` 的浏览器回调流)
-- 之后:主界面显示 agent 活动 + 最近纪要(点开 / 打开文件夹)
+- 之后:主界面显示 agent 活动和最近纪要，可打开产物或重试失败录音
 
 ### 打包内容
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@fastagent-sh/voicenote",
-  "version": "0.22.0",
+  "version": "0.22.1",
   "description": "Voice recordings → diarized transcripts → integrated semantic Markdown notes. Currently optimized for the PHILIPS VTR6500 recorder, but the workflow is generic.",
   "type": "module",
   "license": "MIT",

package/src/cli.ts

@@ -3,7 +3,7 @@ import { cac } from 'cac'
 import packageJson from '../package.json' with { type: 'json' }
 import { parseLockOwner } from './runLock'
 import { tosObject, type TosConfig as VolcanoTosConfig } from './tos'
-import { applyOutcome, buildJobsView, classify, emptyState, localIso, MAX_ATTEMPTS, migrateLegacyState, ownsOutput, parseJobsLimit, parseStateFile, parseStrictJson, patchJob, pruneUnseen, reconcileInterrupted, startAttempt, SUMMARY_FAILED_STATUS, type CurrentJob, type JobRecord, type StateFile } from './jobs'
+import { applyOutcome, buildJobsView, classify, emptyState, localIso, MAX_ATTEMPTS, migrateLegacyState, ownsOutput, parseJobsLimit, parseStateFile, parseStrictJson, patchJob, pruneUnseen, reconcileInterrupted, requeueFailed, startAttempt, SUMMARY_FAILED_STATUS, type CurrentJob, type JobRecord, type StateFile } from './jobs'
 import { createHash, randomUUID } from 'node:crypto'
 import { appendFile, chmod, mkdir, readFile, writeFile, copyFile, rename, unlink, stat, readdir } from 'node:fs/promises'
 import { existsSync, readFileSync, readdirSync, mkdirSync, writeFileSync, appendFileSync, openSync, closeSync, statSync, readSync, unlinkSync, renameSync } from 'node:fs'
@@ -2310,6 +2310,21 @@ async function forgetRecording(needle: string): Promise<void> {
   } finally { await lock.release() }
 }
 
+async function retryRecording(id: string): Promise<void> {
+  const config = getConfig()
+  const lock = await acquireRunLock()
+  if (!lock) throw new Error('A voicenote run is in progress. Retry once it finishes.')
+  try {
+    await migrateStateOnDisk(config)
+    const store = await loadState(config)
+    const entry = store.jobs[id]
+    if (!entry) throw new Error('Recording no longer exists in the processing list.')
+    if (!requeueFailed(entry, nowIso())) throw new Error(`Cannot retry a recording in state '${entry.state}'.`)
+    await saveState(config, store)
+    console.log(`queued ${entry.name} for retry`)
+  } finally { await lock.release() }
+}
+
 async function showLog(opts: { lines?: number; follow?: boolean; err?: boolean; date?: string }): Promise<void> {
   const lines = Number(opts.lines || 30)
   const wanted = [opts.date ? join(LOG_DIR, `${opts.date}.log`) : dailyLogPath()]
@@ -2627,6 +2642,7 @@ cli.command('jobs', 'Show every recording\'s processing status (running, queued,
 cli.command('open [target]', 'Open notes dir, config dir (`config`), logs dir (`logs`), or a note matching the slug').action((target?: string) => openTarget(target))
 
 cli.command('forget <key>', 'Drop a recording\'s job record so it is queued again (a saved transcript on disk is still reused)').action((key: string) => forgetRecording(key))
+cli.command('retry <id>', 'Requeue one failed recording while retaining saved outputs').action((id: string) => retryRecording(id))
 
 cli.command('log', 'Print the daily log (today by default)')
   .option('--lines <n>', 'How many trailing lines to print', { default: 30 })

package/src/jobs.ts

@@ -163,6 +163,13 @@ export function startAttempt(entry: JobRecord, now: string): void {
   patchJob(entry, { state: 'running', code: null, detail: null, attempts: entry.attempts + 1 }, now)
 }
 
+/** Manually requeue a failed job while retaining any saved transcript/audio. */
+export function requeueFailed(entry: JobRecord, now: string): boolean {
+  if (entry.state !== 'error' && entry.state !== 'gave_up') return false
+  patchJob(entry, { state: 'queued', detail: null, attempts: 0 }, now)
+  return true
+}
+
 /**
  * Reclaim records left `running` by a dead run. Safe to do wholesale because the
  * caller holds the run lock: no other run can own a `running` record right now.
@@ -241,6 +248,7 @@ export function pruneUnseen(jobs: Record<string, JobRecord>, seen: Set<string>,
 export type CurrentJob = { pid: number; source_id: string; step: string; started_at: string }
 
 type JobView = {
+  id: string | null
   status: 'running' | 'queued' | 'done' | 'notes_failed' | 'error' | 'gave_up' | 'filtered'
   name: string
   title: string | null
@@ -356,6 +364,7 @@ function foldFiltered(records: JobRecord[]): JobView | null {
   }
   const detail = [...counts].map(([label, n]) => `${label} ×${n}`).join(', ')
   return {
+    id: null,
     status: 'filtered',
     name: `${records.length} recording${records.length > 1 ? 's' : ''} filtered out`,
     title: null, time: null, step: null, detail, notes: null,
@@ -383,6 +392,7 @@ export function buildJobsView(
 
   for (const [id, j] of Object.entries(state.jobs ?? {})) {
     const base = {
+      id,
       name: j.name,
       title: j.title ?? null,
       time: displayTime(j.recorded_at),