Skip to content

an open-source alternative to the dots, bots and muses of the world, run an office of claude code/codex like agents on your laptop, sandboxes or anywhere, uses your existing subscriptions

docs: v0.3.1 = three CLI engines (OpenCode/Crush/pi.dev) + BYOK/local + two fixes

Chaitanya Giri5590d19e9461
diff --git a/CHANGELOG.md b/CHANGELOG.md
index ada322da..9c7ed911 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -6,9 +6,34 @@ All notable changes to this project are documented here. The format is based on
## [0.3.1] — 2026-06-22
-A reliability patch for the hive message router. Completes the **wake-reliability hardening**
-started in 0.3.0 (which caught up missed *schedules* and auto-revived wedged *terminals* on wake,
-but left the *router* out).
+Three more coding CLIs join the floor — **OpenCode**, **Crush**, and **pi.dev** — each usable as a
+worker *and* as Michael, with **bring-your-own keys + local LLMs**. Plus two reliability fixes: the
+sleep-frozen message router and Codex workers' filesystem permissions.
+
+> **Live verification note:** the three new engines are wired end-to-end and selectable as god, but
+> their bridges' *runtime* behavior (real model calls → tool gating → turn-end drain) is **pending
+> BYOK keys / a local LLM** — verify on-device. Their architecture (preset + bridge + payload
+> contract) was reviewed line-by-line; the guaranteed mail-drain path (the renderer idle inbox-wake
+> nudge) is provider-agnostic and already proven.
+
+### Added
+- **Three new selectable engines: OpenCode · Crush · pi.dev.** Each lands as a declarative
+ `AgentProviderPreset` and appears automatically in the Add-Agent picker (worker) and the god
+ engine picker (orchestrator). Each gets a **bridge** for live status + turn-end inbox-drain:
+
+ | Engine | Identity | Bridge | Notes |
+ |---|---|---|---|
+ | **OpenCode** | `opencode` (anomalyco/opencode, TS) | **native plugin** (`session.idle`) | bundled per-agent plugin; no traffic interception; auto-approve via gated `OPENCODE_CONFIG_CONTENT` |
+ | **Crush** | `crush` (charmbracelet, Go) | **proxy** (qwen-tier) | per-agent `CRUSH_GLOBAL_CONFIG` routes traffic through the loopback sidecar (Crush has no base-URL env) |
+ | **pi.dev** | `pi` (earendil-works) | **hooks** (bundled extension) | `pi.on(event)` → HIVE_SOCK; extension auto-approves tools only when the floor is in auto mode |
+
+- **BYOK + local-LLM config UI (Settings → AI Engines).** A new per-provider config surface: API
+ keys for the backend model-providers (Anthropic / OpenAI / Google / OpenRouter / Groq) stored
+ **write-only** in the encrypted secret broker (never read back to the renderer; materialized
+ main-only at spawn), plus per-engine **local base-URL** + default-model fields
+ (`HarnessConfig.providerBaseUrls` / `providerDefaultModels`). Pi/OpenCode/Crush/Qwen pick up the
+ keys + endpoints at spawn; auto-mode stays gated behind the floor toggle, and each engine runs
+ unsandboxed in auto mode (surfaced as a caveat).
### Fixed
- **Codex hive workers get full filesystem + auto-approval from spawn (parity with Claude).** A Codex-engine agent in auto mode launched with `-a never -s workspace-write`, whose sandbox scopes writes to the PTY cwd (the user's project). But a hive worker must also write to its agent folder at `<harnessHome>/hive/agents/<id>/` (move `inbox/` → `.done/`, append `memory.md`, drop outbox JSON, write deliverables) — a **different path tree from cwd**, which `workspace-write` blocked. So a freshly spawned Codex worker couldn't complete HIVE PROTOCOL housekeeping and reported "it does not have permissions … grant write permission to the agent folder." Codex's auto-mode flag is now `--dangerously-bypass-approvals-and-sandbox` — the documented equivalent of Claude's `bypassPermissions` / Antigravity's `--dangerously-skip-permissions` (skip all approval prompts **and** drop the OS sandbox), so a Codex worker has the same filesystem access and auto-approval as a Claude worker from the get-go (`src/shared/agentProvider.ts`; reference/copy updated in `src/shared/codexCommands.ts`, `OnboardingWizard.tsx`, `renderer/store/config.ts`). Claude/agy/antigravity behavior is unchanged.
diff --git a/README.md b/README.md
index 1ac37572..d2e1c2ee 100644
--- a/README.md
+++ b/README.md
@@ -117,7 +117,7 @@ terminal/event plane, and [`DESIGN.md`](./DESIGN.md) for the visual system.
| **MemoryReflector** | Memory condensation that summarizes and bounds per-agent memory over time, so long-term memory doesn't grow without limit. |
| **Configurable home folder** | Point the hive/memory home at any folder, with a safe move that relocates existing state without losing it. |
| **Restore team** | One-click "Restore team" rebuilds last session's workers after a harness restart — no more re-adding agents by hand. |
-| **Selectable agent engines** | Each agent — and Michael himself — runs on a pluggable engine: Claude Code, Antigravity, OpenAI Codex, or a local provider (a claw/qwen backend proxy). Choose the engine per hire from a visual provider/hive picker; the orchestrator's own engine is swappable from onboarding or a change-engine flow. |
+| **Selectable agent engines** | Each agent — and Michael himself — runs on a pluggable engine: Claude Code, Antigravity, OpenAI Codex, **OpenCode**, **Crush**, **pi.dev**, or a local provider (a claw/qwen backend proxy). Choose the engine per hire from a visual provider/hive picker; the orchestrator's own engine is swappable from onboarding or a change-engine flow. BYOK keys + local-LLM endpoints for the CLI engines live in **Settings → AI Engines**. |
| **Per-hire skills + MCP catalog** | Every hire carries a manifest of allowed skills + MCP servers (default-deny over a shared catalog). Bundled skills ship with the app, and a consent UI surfaces every skill/MCP a hire wants before it can use it — untrusted hire input is reviewed, never auto-granted. |
| **Integrations registry + secret broker** | A declarative integrations registry with a registry-driven Settings UI and a loopback secret broker: secrets are write-only (set once, never read back into the renderer) and reached only through the broker. Ships with a first wave of declarative templates. |
| **Slack-spawned ephemeral workers** | Michael can spawn an isolated worker straight from a Slack request, have it post its reply back into the thread, then tear it down safely — with worktree GC, per-worker token caps, and a teardown gate that never auto-discards unintegrated work. Live workers appear in a dedicated Workers tab. |
@@ -287,7 +287,7 @@ Shipped in **v0.2.0–v0.3.0**:
Next up:
- [ ] **More chat integrations** — Telegram and richer chat bridges that pipe a channel straight into Michael's queue (and route his replies back out), so you can run the floor from your phone.
-- [ ] **More engines & providers** — the engine abstraction is in (Claude Code, Antigravity, Codex, local providers); keep adding engines and broadening the per-hire capability catalog on top of it.
+- [x] **More engines & providers** — the engine abstraction is in (Claude Code, Antigravity, Codex, local providers), and **v0.3.1 adds OpenCode, Crush, and pi.dev** as worker+god engines with BYOK keys + local LLMs (live runtime pending on-device key verification). Keep adding engines and broadening the per-hire capability catalog on top of it.
- [ ] **More integration templates** — grow the integrations registry beyond the first-wave templates.
- [ ] **Realtime Michael** — a low-latency realtime LLM channel for quick, snappy back-and-forth with the orchestrator, alongside the async terminal.
- [ ] **Fuller avatar coverage** — push the remaining station visits and tool-bubbles to be driven 100% by real Claude Code hook events.
diff --git a/RELEASE.md b/RELEASE.md
index efa84f7a..2e1841ca 100644
--- a/RELEASE.md
+++ b/RELEASE.md
@@ -7,15 +7,29 @@ routing, and remembering, coordinated by a GOD orchestrator you talk to. Local-f
---
-## What's new in 0.3.1 — *Router survives sleep*
-
-A reliability patch. The hive message router (the loop that drains each agent's `outbox/` into
-recipients' `inbox/`) is a timer that **freezes during true macOS system sleep**. 0.3.0's
-wake-reliability hardening re-armed the scheduler and revived wedged terminals on wake, but the
-router was left out — so after a long sleep (laptop closed overnight) the GOD orchestrator's
-standups kept arriving while **agent-to-agent and god-to-worker mail silently stopped delivering**.
-On wake the harness now **re-arms the router and immediately drains whatever piled up**, so
-delegation keeps flowing after the machine sleeps. (Verified by `scripts/verify-keepalive-catchup.mjs`.)
+## What's new in 0.3.1 — *Three more engines: OpenCode · Crush · pi.dev*
+
+The floor gets three new coding CLIs, each usable as a **worker and as Michael**, with
+**bring-your-own keys + local LLMs**:
+
+- **OpenCode** (`opencode`, the TypeScript agent) — wired via a **native plugin** bridge
+ (`session.idle`): no traffic interception, provider-agnostic, works with local models.
+- **Crush** (`crush`, Charmbracelet's Go TUI) — a **proxy** bridge routes its LLM traffic through a
+ per-agent config so a hookless CLI still reports status and drains mail.
+- **pi.dev** (`pi`, the Pi Coding Agent) — a **hooks** bridge: a bundled extension maps Pi's
+ `pi.on(event)` lifecycle onto the hive, and auto-approves tools only when the floor is in auto mode.
+
+A new **Settings → AI Engines** panel collects per-provider **API keys** (stored *write-only*,
+encrypted, never shown again) and **local base-URLs** (Ollama / LM Studio / vLLM) for these engines.
+
+> **Heads-up:** these engines are integrated and selectable as god, but **live end-to-end
+> verification with real model calls is pending your BYOK keys / a local LLM** — verify on-device.
+> The guaranteed inbox-drain path (the idle wake-nudge) is provider-agnostic and already proven.
+
+Plus two reliability fixes: the **message router now survives system sleep** (it re-arms and drains
+the backlog on wake, so god→worker and agent↔agent mail keeps flowing after the laptop's been
+closed), and **Codex workers get full filesystem + auto-approval from spawn** (parity with Claude),
+so a fresh Codex hive worker can do its protocol housekeeping without a permission prompt.
---
diff --git a/docs/index.html b/docs/index.html
index 003246b0..3fe08968 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -902,12 +902,15 @@
<div class="fb-ic">🔌</div>
<div class="fb-content">
<div class="fb-tag">Multi-CLI · Full parity</div>
- <div class="fb-head">Three CLIs. One hive.</div>
- <div class="fb-desc">Claude Code, Antigravity (Gemini), and OpenAI Codex each get a desk, a mailbox, and full hive integration. Mix and match — the right capability on the right task.</div>
+ <div class="fb-head">Six CLIs. One hive.</div>
+ <div class="fb-desc">Claude Code, Antigravity (Gemini), OpenAI Codex, and now <b>OpenCode</b>, <b>Crush</b> &amp; <b>pi.dev</b> — each gets a desk, a mailbox, and full hive integration, as a worker or as Michael. Bring your own keys + local LLMs (Settings → AI Engines). Mix and match — the right capability on the right task.</div>
<div class="fb-badges">
<span class="fb-badge"><span class="dot" style="background:var(--sky);"></span>Claude Code</span>
<span class="fb-badge"><span class="dot" style="background:var(--mint);"></span>Antigravity</span>
<span class="fb-badge"><span class="dot" style="background:var(--peach);"></span>Codex</span>
+ <span class="fb-badge"><span class="dot" style="background:var(--lilac);"></span>OpenCode</span>
+ <span class="fb-badge"><span class="dot" style="background:var(--coral);"></span>Crush</span>
+ <span class="fb-badge"><span class="dot" style="background:var(--yellow);"></span>pi.dev</span>
</div>
</div>
</div>