feat(fleet): comms-block emitter + FLEET-LAUNCH runbook (#633)
Add `mosaic fleet comms-block <role> [--host]` — an explicit-arg, comms-only emitter wrapping a new resolveCommsBlock() over readFleetCommsBlock. Unlike the launch-time reader (which returns '' on any miss so composeContract can no-op silently), the emitter fails loud: unknown role / missing roster → stderr + exit 1, so an operator can safely preview any peer's view and a typo is never a silent no-op. Add docs/fleet/FLEET-LAUNCH.md: the canonical roster-driven launch path (worker + orchestrator .env fold via MOSAIC_AGENT_COMMAND, which short-circuits the line-44 yolo hardcode), 3 launch gotchas (flag conflict, MOSAIC_AGENT_NAME baking, launchRuntime guards), the #632 preserve-list note, and the North-Star A→B→webUI launch-config arc. PATH A of the orchestrator-launch fix; PATH B (roster-native launch-config: yolo toggle + command/channels emission) tracked as #636. TDD: 6 new resolveCommsBlock cases; 177 fleet+comms tests green; typecheck, eslint, prettier clean. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EsgTQzV5YUGk1JtCLP4B83
This commit is contained in:
@@ -86,3 +86,7 @@ Active workstream is **W1 — Federation v1**. Workers should:
|
||||
## #631 — re-seed preserves user fleet data (CRITICAL) — fix/631-reseed-preserves-fleet-data
|
||||
|
||||
- Status: implemented + tested. PRIMARY: install.sh PRESERVE_PATHS += fleet/\*.yaml + fleet/agents + fleet/run (glob-aware cp-fallback); TS parity. SECONDARY: refreshActiveFleetUnits propagates unit fixes to ~/.config/systemd/user on mosaic update. bash F6 + TS + unit tests green. Detail: scratchpads/631-reseed-preserves-fleet.md.
|
||||
|
||||
## #633 — comms-block emitter + FLEET-LAUNCH runbook — feat/633-comms-block-runbook
|
||||
|
||||
- Status: implemented + tested (TDD). `mosaic fleet comms-block <role> [--host]` wraps resolveCommsBlock → readFleetCommsBlock; fails loud (stderr + exit 1) on unknown role / missing roster instead of silent empty. docs/fleet/FLEET-LAUNCH.md runbook: worker path + orchestrator .env fold (MOSAIC_AGENT_COMMAND; line-41 [-z] short-circuits line-44 yolo hardcode) + 3 launch gotchas + #632 preserve note + North-Star 4-field arc (harness ✅/model ✅ roster-native today; yolo + command/channels = PATH B #636). 177 fleet+comms tests green (6 new resolveCommsBlock cases). PATH A of the A→B→webUI arc. Detail: scratchpads/633-comms-block-runbook.md.
|
||||
|
||||
114
docs/fleet/FLEET-LAUNCH.md
Normal file
114
docs/fleet/FLEET-LAUNCH.md
Normal file
@@ -0,0 +1,114 @@
|
||||
# Fleet Launch Runbook
|
||||
|
||||
How every Mosaic fleet agent — workers **and** the orchestrator — is launched, and how to
|
||||
configure each one. The guiding principle: **one roster-driven launcher**. There is no bespoke
|
||||
per-agent launch script; the roster plus per-agent `.env` files are the single source of launch
|
||||
config.
|
||||
|
||||
## The launch chain
|
||||
|
||||
| Layer | File | Responsibility |
|
||||
| ---------------- | ------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------- |
|
||||
| systemd unit | `mosaic-agent@<role>.service` | One templated unit per role; `ExecStart` runs the session launcher with the instance name `%i`. Defaults `MOSAIC_AGENT_RUNTIME=pi`, `MOSAIC_AGENT_NAME=%i`. |
|
||||
| session launcher | `tools/fleet/start-agent-session.sh <role>` | Builds the launch command, opens the tmux pane, wires the heartbeat. |
|
||||
| launch command | `mosaic yolo <runtime>` (or a per-agent override) | Replaces the pane's foreground process with the runtime, fully seeded. |
|
||||
| seeding | `mosaic`'s `composeContract()` | Injects the Constitution/USER/TOOLS/runtime contract, `*.local` overlays, **and** the Fleet-Comms cheat-sheet — all via `--append-system-prompt`. |
|
||||
|
||||
Per-agent overrides live in `fleet/agents/<role>.env`, generated from `roster.yaml` by
|
||||
`generateAgentEnv` (`packages/mosaic/src/commands/fleet.ts`) and consumed by the launcher.
|
||||
|
||||
## Worker launch path (default)
|
||||
|
||||
1. `roster.yaml` carries each agent's `runtime` and optional `model_hint`.
|
||||
2. `generateAgentEnv` emits `fleet/agents/<role>.env` with `MOSAIC_AGENT_NAME`,
|
||||
`MOSAIC_AGENT_RUNTIME`, and `MOSAIC_AGENT_MODEL`.
|
||||
3. `start-agent-session.sh` has no `MOSAIC_AGENT_COMMAND` set, so it falls through to the default
|
||||
(line ~44):
|
||||
```sh
|
||||
MOSAIC_AGENT_COMMAND="mosaic yolo $MOSAIC_AGENT_RUNTIME${MOSAIC_AGENT_MODEL:+ --model $MOSAIC_AGENT_MODEL}"
|
||||
```
|
||||
4. The launcher bakes `MOSAIC_AGENT_NAME` into the pane command (line ~118), so `composeContract`
|
||||
can inject the Fleet-Comms cheat-sheet for that role.
|
||||
|
||||
That is the whole worker path: roster → `.env` → `mosaic yolo <runtime>` → seeded pane.
|
||||
|
||||
## Orchestrator fold (PATH A — ships today)
|
||||
|
||||
The orchestrator is **just another roster agent** launched through the canonical path — not a
|
||||
snowflake script.
|
||||
|
||||
| Piece | Value |
|
||||
| ------------------ | ----------------------------------- |
|
||||
| host-side launcher | `orchestrator-launch.sh` |
|
||||
| systemd unit | `mosaic-fleet-orchestrator.service` |
|
||||
| tmux session | `orchestrator` (role-named) |
|
||||
|
||||
Set its launch command via `fleet/agents/orchestrator.env`:
|
||||
|
||||
```sh
|
||||
MOSAIC_AGENT_COMMAND='mosaic yolo claude --channels plugin:discord@<channel>'
|
||||
```
|
||||
|
||||
When `MOSAIC_AGENT_COMMAND` is set, `start-agent-session.sh`'s `if [ -z "$MOSAIC_AGENT_COMMAND" ]`
|
||||
guard (line ~41) is false, so the line-44 default — **including its hardcoded `yolo`** — is skipped
|
||||
entirely. The override fully controls the runtime and flags. Routing through `mosaic yolo claude`
|
||||
(rather than a raw `claude` invocation) is what gives the orchestrator the same full
|
||||
`composeContract` seeding + Fleet-Comms cheat-sheet as every worker, with `--channels` and any
|
||||
other flags passed straight through to the `claude` binary.
|
||||
|
||||
## Launch gotchas
|
||||
|
||||
1. **Flag conflict.** `mosaic yolo claude` already injects `--dangerously-skip-permissions`. Do
|
||||
**not** also pass `--permission-mode bypassPermissions` — the `claude` binary would receive both.
|
||||
Use `mosaic yolo claude …` alone (yolo covers the unattended posture), **or** non-yolo
|
||||
`mosaic claude --permission-mode bypassPermissions …`. Never mix the two.
|
||||
2. **`MOSAIC_AGENT_NAME` must reach the pane.** The launcher bakes it from the instance name, and
|
||||
`composeContract` gates the Fleet-Comms block on it (`launch.ts`, in `composeContract`) — **and**
|
||||
the role must be a member of `roster.yaml`, or the block resolves empty.
|
||||
3. **`launchRuntime` guards.** `mosaic yolo claude` runs `checkSoul` / `checkRuntime` /
|
||||
`checkSequentialThinking`. The host needs `SOUL.md` and the sequential-thinking MCP, or the
|
||||
launch aborts (a raw `claude` invocation skipped these checks). Dry-run the composed command in a
|
||||
throwaway tmux session before swapping a live launcher.
|
||||
|
||||
## Why per-agent `.env` survives upgrades (#632)
|
||||
|
||||
`install.sh` `PRESERVE_PATHS` includes `fleet/*.yaml`, `fleet/agents`, and `fleet/run`, so
|
||||
`mosaic update`'s framework re-seed **preserves** your roster and per-agent `.env` overrides
|
||||
(glob-aware `cp` fallback; matching TS parity in `file-adapter.ts`). Before #632, an auto re-seed
|
||||
could wipe them — which is exactly why PATH A's `.env` override is safe to rely on now.
|
||||
|
||||
## Inspecting the comms wiring
|
||||
|
||||
- `mosaic fleet comms-block <role>` prints the Fleet-Comms cheat-sheet a given role receives at
|
||||
launch — its `[host:session]` identity, the exact `agent-send.sh` command for each peer, and the
|
||||
FLIP / `--verify` conventions. `--host <h>` previews a cross-host view. An unknown role or missing
|
||||
roster **fails loud** (stderr + non-zero exit), so a typo is never a silent no-op.
|
||||
- Versus `mosaic compose-contract <runtime>`: that emits the **whole** system prompt and reads the
|
||||
role from `MOSAIC_AGENT_NAME` (a full-prompt smoke test). `comms-block` is the targeted,
|
||||
explicit-arg, comms-only view — e.g. `mosaic fleet comms-block coder0-0` to preview a peer.
|
||||
|
||||
## North Star / future direction
|
||||
|
||||
**Vision:** a webUI lets the user edit each agent's launch config — switch **harness**
|
||||
(claude / pi / codex / opencode), toggle **yolo**, pick a **model**, set a **command/channels**
|
||||
override — with no terminal.
|
||||
|
||||
**Continuity — this is not a new launch path.** It is a data-model + UI-binding layer over the
|
||||
existing roster-driven launcher. Field-by-field status today:
|
||||
|
||||
| Launch-config field | Roster-native today? | Mechanism / gap |
|
||||
| ------------------------ | -------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------- |
|
||||
| **harness** (`runtime`) | ✅ end-to-end | `roster.runtime` → `generateAgentEnv` emits `MOSAIC_AGENT_RUNTIME` → launcher line 44. UI just writes the field. |
|
||||
| **model** (`model_hint`) | ✅ end-to-end | `roster.model_hint` → `MOSAIC_AGENT_MODEL` → launcher line 44 `--model`. UI just writes the field. |
|
||||
| **yolo** | ❌ new | Launcher line 44 **hardcodes** `mosaic yolo`. A non-yolo toggle needs a roster `yolo` field → emit `MOSAIC_AGENT_YOLO` → make line 44 conditional. |
|
||||
| **command / channels** | ❌ new | `MOSAIC_AGENT_COMMAND` is **consumed** (launcher line ~12) but `generateAgentEnv` does not emit it. Needs a roster `command`/`channels` field → emitted. |
|
||||
|
||||
**The arc:**
|
||||
|
||||
- **A** — `.env` `MOSAIC_AGENT_COMMAND` hatch: manual, ships now, kept safe across upgrades by #632.
|
||||
- **B** — roster-native launch-config: harness + model are already there; add the **yolo** toggle
|
||||
(line-44 conditional) and **command/channels** emission to complete the data model.
|
||||
- **webUI** — binds dropdowns/toggles directly to those four roster fields.
|
||||
|
||||
PATH A's `.env` override is the **manual form** of exactly what PATH B makes roster-native and the
|
||||
webUI edits — one continuous arc, not three separate features. PATH B is tracked as #636.
|
||||
54
docs/scratchpads/633-comms-block-runbook.md
Normal file
54
docs/scratchpads/633-comms-block-runbook.md
Normal file
@@ -0,0 +1,54 @@
|
||||
# #633 — comms-block emitter + FLEET-LAUNCH runbook
|
||||
|
||||
Branch: `feat/633-comms-block-runbook` (off `bf2a6745`, post-#632 merge)
|
||||
Issue: #633 · Follow-up filed: #636 (PATH B)
|
||||
|
||||
## Goal
|
||||
|
||||
PATH A of the orchestrator-launch fix: give every launch path the Fleet-Comms onboarding, and
|
||||
document the canonical roster-driven launcher so the orchestrator stops being a bespoke snowflake.
|
||||
|
||||
## Deliverables
|
||||
|
||||
1. **`mosaic fleet comms-block <role> [--host <h>]`** — explicit-arg, comms-block-only emitter.
|
||||
- Backed by new `resolveCommsBlock(mosaicHome, role, fleetHost?)` in `fleet/comms-onboarding.ts`
|
||||
returning `{ ok, output, error }`.
|
||||
- Unlike `readFleetCommsBlock` (returns `''` on any miss so `composeContract` can no-op silently
|
||||
during launch), the emitter **fails loud**: unknown role / missing roster → `ok:false` → CLI
|
||||
prints to stderr + sets `process.exitCode = 1`. A typo is never a silent no-op.
|
||||
- Distinct from `mosaic compose-contract <runtime>` (whole prompt, env-coupled via
|
||||
`MOSAIC_AGENT_NAME`); comms-block is the targeted, explicit-arg, comms-only view.
|
||||
2. **`docs/fleet/FLEET-LAUNCH.md`** — worker path + orchestrator `.env` fold + 3 launch gotchas +
|
||||
#632 preserve note + North-Star 4-field arc.
|
||||
|
||||
## Key findings (drove the design)
|
||||
|
||||
- `mosaic yolo claude` **already** forwards `--channels`/`--permission-mode` to the binary
|
||||
(`launch.ts` claude case `cliArgs.push(...args)`) AND injects the comms block via
|
||||
`composeContract` → `readFleetCommsBlock(home, env.MOSAIC_AGENT_NAME)`. So no `launch.ts` change
|
||||
was needed — PATH A is `.env` + doc only.
|
||||
- `start-agent-session.sh` line ~41 `[ -z "$MOSAIC_AGENT_COMMAND" ]` short-circuits the line-44
|
||||
default, so an `.env` `MOSAIC_AGENT_COMMAND` override bypasses the hardcoded `yolo` entirely — the
|
||||
yolo-conditional is therefore a PATH B (default-path) concern, not PATH A.
|
||||
- `generateAgentEnv` (`fleet.ts` ~202-207) emits NAME/RUNTIME/MODEL but **not** `MOSAIC_AGENT_COMMAND`
|
||||
— the seam PATH B (#636) closes.
|
||||
|
||||
## A → B → webUI arc (North Star)
|
||||
|
||||
- A = `.env` `MOSAIC_AGENT_COMMAND` hatch (manual, ships now, #632-safe).
|
||||
- B (#636) = roster-native launch-config: harness ✅ + model ✅ already there; add **yolo** (line-44
|
||||
conditional `MOSAIC_AGENT_YOLO`) + **command/channels** (`generateAgentEnv` emission).
|
||||
- webUI binds dropdowns/toggles to those four roster fields. One launcher, no new launch path.
|
||||
|
||||
## Results
|
||||
|
||||
- TDD: spec first (`comms-onboarding.spec.ts`, 6 new `resolveCommsBlock` cases) → red → implement → green.
|
||||
- `fleet.spec.ts` subcommand-list assertion extended with `comms-block`.
|
||||
- 177 fleet+comms tests green; typecheck clean; eslint clean; prettier clean.
|
||||
|
||||
## Risks / notes
|
||||
|
||||
- Pre-existing local-only failure `uninstall.spec.ts > removeFramework > handles missing mosaicHome
|
||||
gracefully` (EACCES on `/nonexistent` as non-root) — unrelated to #633, passes in CI as root.
|
||||
- Did NOT run `mosaic update` / anything auto-reseed: installed CLI still 0.0.40 (roster-wipe live
|
||||
until mos-claude-0 ships 0.0.41). All work is in-repo + vitest, never touches the live mosaic home.
|
||||
@@ -95,6 +95,7 @@ describe('registerFleetCommand', () => {
|
||||
expect(agent).toBeDefined();
|
||||
expect(agent!.options.map((option) => option.long)).toContain('--list');
|
||||
expect(agent!.commands.map((command) => command.name()).sort()).toEqual([
|
||||
'comms-block',
|
||||
'reset',
|
||||
'roster',
|
||||
'send',
|
||||
|
||||
@@ -7,6 +7,7 @@ import { spawn } from 'node:child_process';
|
||||
import * as readline from 'node:readline';
|
||||
import type { Command } from 'commander';
|
||||
import YAML from 'yaml';
|
||||
import { resolveCommsBlock } from '../fleet/comms-onboarding.js';
|
||||
|
||||
/**
|
||||
* A function that spawns a command with inherited stdio (TTY passthrough).
|
||||
@@ -1359,6 +1360,23 @@ export function registerFleetAgentCommands(
|
||||
}
|
||||
});
|
||||
|
||||
agentCommand
|
||||
.command('comms-block <role>')
|
||||
.description(
|
||||
"Print the Fleet Comms cheat-sheet for a roster role (preview a peer's peer-reach view)",
|
||||
)
|
||||
.option('--host <host>', 'Override the fleet host (preview a cross-host peer view)')
|
||||
.action((role: string, opts: { host?: string }) => {
|
||||
const mosaicHome = resolveMosaicHomeFromCommand(agentCommand, deps.mosaicHome);
|
||||
const res = resolveCommsBlock(mosaicHome, role, opts.host);
|
||||
if (!res.ok) {
|
||||
console.error(`[mosaic] comms-block: ${res.error}`);
|
||||
process.exitCode = 1;
|
||||
return;
|
||||
}
|
||||
console.log(res.output);
|
||||
});
|
||||
|
||||
agentCommand
|
||||
.command('status [agent]')
|
||||
.description('Show tmux status for the local fleet or one agent')
|
||||
|
||||
@@ -7,6 +7,7 @@ import {
|
||||
buildFleetCommsBlock,
|
||||
renderPeerReach,
|
||||
readFleetCommsBlock,
|
||||
resolveCommsBlock,
|
||||
type CommsPeer,
|
||||
} from './comms-onboarding.js';
|
||||
|
||||
@@ -185,3 +186,53 @@ describe('readFleetCommsBlock — situational (the context a spawned agent gets)
|
||||
expect(readFleetCommsBlock(mkdtempSync(join(tmpdir(), 'noroster-')), 'orchestrator')).toBe('');
|
||||
});
|
||||
});
|
||||
|
||||
describe('resolveCommsBlock — `mosaic fleet comms-block <role>` emitter semantics', () => {
|
||||
// The emitter wraps readFleetCommsBlock but must NEVER print an empty string silently:
|
||||
// an unknown role / missing roster has to fail loud (caller maps !ok → stderr + exit 1)
|
||||
// so `mosaic fleet comms-block bogus` is a visible error, not a confusing no-op. The
|
||||
// success path returns the block verbatim for `mosaic fleet comms-block <peer>` previews.
|
||||
let home: string;
|
||||
beforeEach(() => {
|
||||
home = mkdtempSync(join(tmpdir(), 'mosaic-commsblk-'));
|
||||
mkdirSync(join(home, 'fleet'), { recursive: true });
|
||||
writeFileSync(join(home, 'fleet', 'roster.yaml'), ROSTER);
|
||||
});
|
||||
afterEach(() => rmSync(home, { recursive: true, force: true }));
|
||||
|
||||
it('returns ok + the cheat-sheet for a roster member', () => {
|
||||
const res = resolveCommsBlock(home, 'orchestrator', 'w-jarvis');
|
||||
expect(res.ok).toBe(true);
|
||||
expect(res.output).toContain('# Fleet Comms');
|
||||
expect(res.output).toContain('| enhancer |');
|
||||
expect(res.error).toBeUndefined();
|
||||
});
|
||||
|
||||
it('fails loud (not ok + error naming the role) for a non-member — never silently empty', () => {
|
||||
const res = resolveCommsBlock(home, 'stranger', 'w-jarvis');
|
||||
expect(res.ok).toBe(false);
|
||||
expect(res.output).toBe('');
|
||||
expect(res.error).toContain('stranger');
|
||||
});
|
||||
|
||||
it('fails loud when no roster exists at the mosaic home', () => {
|
||||
const noRoster = mkdtempSync(join(tmpdir(), 'mosaic-noroster-'));
|
||||
const res = resolveCommsBlock(noRoster, 'orchestrator', 'w-jarvis');
|
||||
expect(res.ok).toBe(false);
|
||||
expect(res.error).toBeTruthy();
|
||||
rmSync(noRoster, { recursive: true, force: true });
|
||||
});
|
||||
|
||||
it('fails loud for a missing role argument', () => {
|
||||
const res = resolveCommsBlock(home, undefined, 'w-jarvis');
|
||||
expect(res.ok).toBe(false);
|
||||
expect(res.error).toBeTruthy();
|
||||
});
|
||||
|
||||
it('honors a host override so a peer can preview its own cross-host view', () => {
|
||||
// coder0-0 viewing with its own host → its self-identity line uses that host.
|
||||
const res = resolveCommsBlock(home, 'coder0-0', '10.1.10.37');
|
||||
expect(res.ok).toBe(true);
|
||||
expect(res.output).toContain('`[10.1.10.37:coder0-0]`');
|
||||
});
|
||||
});
|
||||
|
||||
@@ -179,5 +179,48 @@ export function readFleetCommsBlock(
|
||||
});
|
||||
}
|
||||
|
||||
/** Result of resolving a comms-block emit request — see `mosaic fleet comms-block`. */
|
||||
export interface CommsBlockResult {
|
||||
/** True when a cheat-sheet was produced; false maps to stderr + non-zero exit. */
|
||||
ok: boolean;
|
||||
/** The Fleet-Comms cheat-sheet (empty unless ok). */
|
||||
output: string;
|
||||
/** Operator-facing reason when !ok. */
|
||||
error?: string;
|
||||
}
|
||||
|
||||
/**
|
||||
* Resolve the Fleet-Comms cheat-sheet for an explicit <role>, backing the
|
||||
* `mosaic fleet comms-block <role>` command. Unlike readFleetCommsBlock — which
|
||||
* returns '' on any miss so composeContract can no-op silently during a launch —
|
||||
* this NEVER silently emits empty: an unknown role or missing roster yields
|
||||
* ok:false + an operator-facing reason, so the CLI surfaces it (stderr + exit 1)
|
||||
* rather than printing nothing. That makes it safe to preview any peer's view,
|
||||
* e.g. `mosaic fleet comms-block coder0-0`.
|
||||
*/
|
||||
export function resolveCommsBlock(
|
||||
mosaicHome: string,
|
||||
role: string | undefined,
|
||||
fleetHost?: string,
|
||||
): CommsBlockResult {
|
||||
if (!role) {
|
||||
return { ok: false, output: '', error: 'comms-block requires a <role> argument' };
|
||||
}
|
||||
const block = fleetHost
|
||||
? readFleetCommsBlock(mosaicHome, role, fleetHost)
|
||||
: readFleetCommsBlock(mosaicHome, role);
|
||||
if (!block) {
|
||||
const rosterPath = join(mosaicHome, 'fleet', 'roster.yaml');
|
||||
return {
|
||||
ok: false,
|
||||
output: '',
|
||||
error: existsSync(rosterPath)
|
||||
? `role "${role}" is not a member of the fleet roster at ${rosterPath}`
|
||||
: `no fleet roster at ${rosterPath}`,
|
||||
};
|
||||
}
|
||||
return { ok: true, output: block };
|
||||
}
|
||||
|
||||
/** Default mosaic home (mirrors launch.ts), for callers that don't pass one. */
|
||||
export const DEFAULT_MOSAIC_HOME_FOR_COMMS = join(homedir(), '.config', 'mosaic');
|
||||
|
||||
Reference in New Issue
Block a user