ai-whisper 0.1.0 → 0.1.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (3) hide show
  1. package/NOTICE +4 -0
  2. package/README.md +122 -0
  3. package/package.json +9 -6
package/NOTICE ADDED
@@ -0,0 +1,4 @@
1
+ ai-whisper
2
+ Copyright 2026 Vu Phan
3
+
4
+ This product is licensed under the Apache License, Version 2.0 (see LICENSE).
package/README.md ADDED
@@ -0,0 +1,122 @@
1
+ # ai-whisper
2
+
3
+ ai-whisper turns two coding agents — Claude and Codex — into a terminal-native pair that hand work back and forth under a single baton, so one agent implements while the other reviews, and a structured workflow drives the loop to a finished, reviewed deliverable without a human babysitting every round.
4
+
5
+ ## Magic moment
6
+
7
+ Mount each agent in its own terminal. Each `mount` claims the current shell, launches the real provider CLI, and binds it to the collab:
8
+
9
+ ```bash
10
+ # terminal 1
11
+ whisper collab mount claude
12
+ # terminal 2
13
+ whisper collab mount codex
14
+ ```
15
+
16
+ Then, from inside either agent's session, kick off a structured workflow against a spec — just ask in plain language:
17
+
18
+ ```text
19
+ Run spec-driven-development using docs/spec.md
20
+ ```
21
+
22
+ From there ai-whisper runs the workflow autonomously:
23
+
24
+ - **Implementer / reviewer assignment** — one agent is the implementer, the other the reviewer (for `spec-driven-development` the default is implementer = Claude, reviewer = Codex). The baton passes between them; only one owns the turn at a time.
25
+ - **Autonomous execution** — the implementer does each step in its real session and hands the result back. An LLM evaluator judges whether the deliverable meets the request.
26
+ - **Review loops** — when work isn't good enough yet, the reviewer's findings are composed into a follow-up handoff and the implementer iterates. The loop repeats until the work is approved or the round budget is exhausted.
27
+ - **Resumability** — workflow and chain state is durable. If the broker restarts or you stop for the day, you recover and reconnect rather than starting over.
28
+ - **Deliverables** — you get committed code plus a review trail (per-step verdicts, round counts), inspectable at any time with `whisper collab dashboard`.
29
+
30
+ ## Visual proof
31
+
32
+ A real `spec-driven-development` run: Claude (left) and Codex (middle) work in their own mounted
33
+ sessions while the dashboard (right) tracks the baton handoffs and per-phase verdicts (~20s).
34
+ Click the still to watch it play on the project page.
35
+
36
+ [![ai-whisper running a spec-driven-development workflow across two mounted agent sessions and the dashboard](docs/assets/workflow-demo-poster.png)](https://ai-creed.dev/projects/ai-whisper/)
37
+
38
+ ## Who this is for
39
+
40
+ ai-whisper is for engineers who already lean on coding agents and want more structure around them:
41
+
42
+ - you already use coding agents heavily and want two of them to check each other.
43
+ - you work terminal-first and want the agents to live in real terminal sessions, not a web UI.
44
+ - you want multi-agent review — a second model gating the first model's output.
45
+ - you run long, structured workflows (spec → plan → implement → review) rather than one-off prompts.
46
+
47
+ It is **not** for:
48
+
49
+ - one-shot "vibe coding" where you just want a quick answer.
50
+ - invisible background automation you never watch.
51
+ - people new to coding agents looking for a guided, hand-holding experience.
52
+
53
+ ## Quickstart
54
+
55
+ Install from npm:
56
+
57
+ ```bash
58
+ npm install -g ai-whisper
59
+ ```
60
+
61
+ Or from a repo checkout:
62
+
63
+ ```bash
64
+ pnpm install
65
+ pnpm build
66
+ ```
67
+
68
+ Install the bundled agent skills once (they let the agents verify, kick off, and report on workflows):
69
+
70
+ ```bash
71
+ whisper skill install
72
+ ```
73
+
74
+ Workflows require an LLM evaluator with credentials — set this up before running one. See [Evaluator configuration](docs/evaluator-configuration.md).
75
+
76
+ Then mount both agents and run a workflow:
77
+
78
+ ```bash
79
+ # terminal 1
80
+ whisper collab mount claude
81
+ # terminal 2
82
+ whisper collab mount codex
83
+ ```
84
+
85
+ The first `mount` creates the collab and starts the broker daemon for the workspace; the second binds the other agent. From either session, start a workflow against a spec or goal file (`spec-driven-development` for a spec, `ralph-loop` for an open-ended goal). Watch it run with:
86
+
87
+ ```bash
88
+ whisper collab dashboard
89
+ ```
90
+
91
+ > Running from a repo checkout instead of a packaged install? Build first (`pnpm build`) and invoke the CLI as `node packages/cli/dist/bin/whisper.js ...` wherever these examples say `whisper ...`.
92
+
93
+ ## Core concepts
94
+
95
+ ai-whisper is **not a swarm**. The agents never type at once — work moves by a single baton, one owner at a time. Mounted sessions are *real* Claude and Codex sessions in your terminal, and those sessions are the source of truth. Autonomy is supervised: every handoff, verdict, and round is inspectable, and runs are resumable rather than fire-and-forget. Work is organized as structured workflows — explicit loops and state transitions, not a free-form chat.
96
+
97
+ Claude and Codex are supported today; the architecture is provider-agnostic by design, so other coding-agent CLIs can be added behind the same relay.
98
+
99
+ For the full mental model, read [Concepts](docs/concepts.md).
100
+
101
+ ## Learn more
102
+
103
+ - [Concepts](docs/concepts.md) — the mental model: baton handoff, real mounted sessions, supervised autonomy, workflow-first execution.
104
+ - [Relay & handoff flows](docs/relay-handoff-flows.md) — the complete handoff state machine, capture-status table, hotkey reference, per-step verdicts, and troubleshooting.
105
+ - [Evaluator configuration](docs/evaluator-configuration.md) — required credentials and options for the LLM evaluator that gates workflows.
106
+ - [Legacy attach mode](docs/legacy-attach.md) — the shelved `attach` / `adopt` flows, kept for historical reference.
107
+
108
+ ## Workspace commands
109
+
110
+ ```bash
111
+ pnpm install
112
+ pnpm test
113
+ pnpm typecheck
114
+ pnpm lint
115
+ pnpm format
116
+ ```
117
+
118
+ ## License
119
+
120
+ Apache License 2.0 — see [LICENSE](LICENSE) and [NOTICE](NOTICE). Contributions are
121
+ accepted under the [Developer Certificate of Origin](CONTRIBUTING.md) (sign off with
122
+ `git commit -s`).
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "ai-whisper",
3
- "version": "0.1.0",
3
+ "version": "0.1.1",
4
4
  "description": "Terminal-first relay for paired AI coding agents (Claude + Codex), driven by structured workflows.",
5
5
  "license": "Apache-2.0",
6
6
  "repository": {
@@ -29,18 +29,21 @@
29
29
  "@types/better-sqlite3": "^7.6.12",
30
30
  "@types/react": "^19.2.14",
31
31
  "ink-testing-library": "^4.0.0",
32
- "@ai-whisper/adapter-claude": "0.0.0",
33
32
  "@ai-whisper/adapter-codex": "0.0.0",
34
- "@ai-whisper/companion-core": "0.0.0",
33
+ "@ai-whisper/broker": "0.0.0",
35
34
  "@ai-whisper/shared": "0.0.0",
36
- "@ai-whisper/broker": "0.0.0"
35
+ "@ai-whisper/adapter-claude": "0.0.0",
36
+ "@ai-whisper/companion-core": "0.0.0"
37
37
  },
38
38
  "files": [
39
39
  "dist",
40
- "skills"
40
+ "skills",
41
+ "README.md",
42
+ "LICENSE",
43
+ "NOTICE"
41
44
  ],
42
45
  "scripts": {
43
- "build": "node scripts/bundle.mjs && node scripts/copy-skills.mjs",
46
+ "build": "node scripts/bundle.mjs && node scripts/copy-skills.mjs && node scripts/copy-package-files.mjs",
44
47
  "typecheck": "tsc -b --emitDeclarationOnly"
45
48
  }
46
49
  }