feat(fleet): onboarding-injection — comms cheat-sheet + peer roster per agent (#621)

Co-authored-by: Jason Woltje <jason@diversecanvas.com>
Co-committed-by: Jason Woltje <jason@diversecanvas.com>

docs/TASKS.md

@@ -70,3 +70,7 @@ Active workstream is **W1 — Federation v1**. Workers should:
 ## F4 — Orchestrator chat connector + Matrix (#616) — feat/f4-matrix-connector
 
 - Status: Phase 1 MERGED (#617: connector interface send/subscribe/health + registry + roster schema + design). Phase 2a (#618): Matrix CS-API client + factory. 20 connector tests green; no fleet.ts changes. Remaining Phase 2: init/configure connector-selection UX + roster wiring, systemd launch wiring, Conduit deploy guide. Detail: scratchpads/f4-matrix-connector.md.
+
+## Fleet onboarding-injection — comms cheat-sheet + peer roster (#620) — feat/fleet-comms-onboarding
+
+- Status: implemented + tested. Injects # Fleet Comms (peer roster + cross-host agent-send commands + FLIP-reply + --verify) into each spawned fleet agent via composeContract; optional per-agent host/ssh/socket roster fields (socket: named → -L, unset → default socket no -L). 10 + 2 tests green. Detail: scratchpads/fleet-comms-onboarding.md.

docs/scratchpads/fleet-comms-onboarding.md

@@ -0,0 +1,31 @@
+# Fleet onboarding-injection — comms cheat-sheet + peer roster (#620)
+
+- **Issue:** #620 · **Branch:** `feat/fleet-comms-onboarding` (off main). Root cause of Mos's failed first send.
+
+## What
+
+Inject a `# Fleet Comms` block into each spawned fleet agent's system prompt (via composeContract — the
+runtime-agnostic path every `mosaic yolo <runtime>` agent hits), so it boots knowing how to reach peers.
+
+- `src/fleet/comms-onboarding.ts` (standalone, no fleet.ts coupling):
+  - `parseRosterAgents` (name/class/host/ssh, lenient), `renderPeerReach` (same-host `-s` vs cross-host
+    `-H <ssh> -s`), `buildFleetCommsBlock` (self [host:session] identity + agent-send path + peer table +
+    FLIP-to-reply + `agent send --verify`=ACCEPTED), `readFleetCommsBlock` (reads roster.yaml; '' if not a member).
+  - `composeContract` appends it only when MOSAIC_AGENT_NAME is set + the agent is in the roster.
+- `roster.schema.json`: optional per-agent `host` + `ssh` (cross-host addresses; manual = pre-federation
+  stopgap, federation/W1 auto-discovers later).
+
+## Acceptance criteria (Mos) — all covered
+
+1. own [host:session] + agent-send path + peer roster ✓
+2. cross-host correctness: local→`-s` (no -H); remote→`-H <ssh> -s` ✓ (concrete coder0-0@dragon-lin)
+3. FLIP-the-preamble reply rule ✓
+4. `agent send --verify` = ACCEPTED ✓
+5. no `-L` (default socket); matches live tooling ✓
+
+## Verification
+
+- 10 onboarding unit tests (parse, render local/remote/fallback/equal-host, build, situational read) +
+  2 composeContract situational tests (injects for fleet agent w/ correct cross-host addr; no-op when
+  MOSAIC_AGENT_NAME unset). tsc/eslint/prettier/sanitize clean.
+- Post-merge validation: Mos spawns a real w-jarvis agent → first-try reach to coder0-0@dragon-lin + a local peer.

packages/mosaic/framework/fleet/roster.schema.json

@@ -81,6 +81,18 @@
           "class": {
             "type": "string"
           },
+          "host": {
+            "description": "Host the agent runs on (hostname or IP). Absent = the fleet host. Used by onboarding-injection to render cross-host comms addresses. Manual cross-host listing is a pre-federation stopgap; federation (W1) auto-discovers later.",
+            "type": "string"
+          },
+          "ssh": {
+            "description": "SSH target (user@host) for a cross-host peer, so onboarding renders the `agent-send.sh -H <user@host>` form. Optional; only needed for agents on a different host than the fleet.",
+            "type": "string"
+          },
+          "socket": {
+            "description": "tmux socket the agent's session runs on. Onboarding renders `-L <socket>` when set; absent = the default socket (no `-L`). Must match the LIVE socket, not blindly inherit the roster's tmux.socket_name.",
+            "type": "string"
+          },
           "working_directory": {
             "type": "string"
           },

packages/mosaic/src/commands/compose-contract.spec.ts

@@ -56,6 +56,55 @@ describe('composeContract — overlay composer', () => {
     rmSync(cwdDir, { recursive: true, force: true });
   });
 
+  it('injects the fleet comms cheat-sheet for a spawned fleet agent (situational)', () => {
+    // A spawned agent has MOSAIC_AGENT_NAME set + is a member of the roster.
+    mkdirSync(join(fixture.home, 'fleet'), { recursive: true });
+    writeFileSync(
+      join(fixture.home, 'fleet', 'roster.yaml'),
+      [
+        'version: 1',
+        'transport: tmux',
+        'agents:',
+        '  - name: orchestrator',
+        '    runtime: claude',
+        '    class: orchestrator',
+        '  - name: enhancer',
+        '    runtime: claude',
+        '    class: enhancer',
+        '  - name: coder0-0',
+        '    runtime: claude',
+        '    class: implementer',
+        '    host: 10.1.10.37',
+        '    ssh: jwoltje@10.1.10.37',
+        '',
+      ].join('\n'),
+    );
+    const prev = process.env['MOSAIC_AGENT_NAME'];
+    try {
+      process.env['MOSAIC_AGENT_NAME'] = 'enhancer';
+      const out = composeContract('claude', fixture.home);
+      expect(out).toContain('# Fleet Comms');
+      expect(out).toMatch(/`\[[^\]]+:enhancer\]`/); // own [host:session] identity (host machine-dependent)
+      // local peer → no -H; cross-host peer → -H ssh
+      expect(out).toContain('-s orchestrator -m "…"');
+      expect(out).toContain('-H jwoltje@10.1.10.37 -s coder0-0 -m "…"');
+      expect(out).not.toContain('-H jwoltje@10.1.10.37 -s orchestrator'); // local stays local
+    } finally {
+      if (prev === undefined) delete process.env['MOSAIC_AGENT_NAME'];
+      else process.env['MOSAIC_AGENT_NAME'] = prev;
+    }
+  });
+
+  it('does NOT inject fleet comms when MOSAIC_AGENT_NAME is unset (non-fleet launch)', () => {
+    const prev = process.env['MOSAIC_AGENT_NAME'];
+    try {
+      delete process.env['MOSAIC_AGENT_NAME'];
+      expect(composeContract('claude', fixture.home)).not.toContain('# Fleet Comms');
+    } finally {
+      if (prev !== undefined) process.env['MOSAIC_AGENT_NAME'] = prev;
+    }
+  });
+
   it('includes the per-tier anchors and the selected harness runtime', () => {
     const out = composeContract('claude', fixture.home);
     expect(out).toContain('GATE-1: the non-negotiable law.'); // L0

packages/mosaic/src/commands/launch.ts

@@ -19,6 +19,7 @@ import { createRequire } from 'node:module';
 import { homedir } from 'node:os';
 import { join, dirname } from 'node:path';
 import type { Command } from 'commander';
+import { readFleetCommsBlock } from '../fleet/comms-onboarding.js';
 
 const MOSAIC_HOME = process.env['MOSAIC_HOME'] ?? join(homedir(), '.config', 'mosaic');
 
@@ -383,6 +384,12 @@ For required push/merge/issue-close/release actions, execute without routine con
   // Runtime-specific contract
   parts.push('\n\n# Runtime-Specific Contract\n\n' + readFileSync(runtimeFile, 'utf-8'));
 
+  // Fleet onboarding: when this is a spawned fleet agent (MOSAIC_AGENT_NAME set
+  // and present in the roster), inject a comms cheat-sheet + peer roster so it
+  // knows how to reach the orchestrator and its peers from its first turn.
+  const fleetComms = readFleetCommsBlock(mosaicHome, process.env['MOSAIC_AGENT_NAME']);
+  if (fleetComms) parts.push('\n\n' + fleetComms);
+
   return parts.join('\n');
 }
 

packages/mosaic/src/fleet/comms-onboarding.spec.ts

@@ -0,0 +1,187 @@
+import { describe, it, expect, beforeEach, afterEach } from 'vitest';
+import { mkdtempSync, mkdirSync, writeFileSync, rmSync } from 'node:fs';
+import { tmpdir } from 'node:os';
+import { join } from 'node:path';
+import {
+  parseRosterAgents,
+  buildFleetCommsBlock,
+  renderPeerReach,
+  readFleetCommsBlock,
+  type CommsPeer,
+} from './comms-onboarding.js';
+
+const ROSTER = [
+  'version: 1',
+  'transport: tmux',
+  'agents:',
+  '  - name: orchestrator',
+  '    runtime: claude',
+  '    class: orchestrator',
+  '  - name: enhancer',
+  '    runtime: claude',
+  '    class: enhancer',
+  '  - name: coder0',
+  '    runtime: pi',
+  '    class: implementer',
+  '  # a manually-listed cross-host peer (pre-federation stopgap)',
+  '  - name: coder0-0',
+  '    runtime: claude',
+  '    class: implementer',
+  '    host: 10.1.10.37',
+  '    ssh: jwoltje@10.1.10.37',
+  '',
+].join('\n');
+
+describe('parseRosterAgents', () => {
+  it('parses name + class + optional host/ssh', () => {
+    const peers = parseRosterAgents(ROSTER);
+    expect(peers.map((p) => p.name)).toEqual(['orchestrator', 'enhancer', 'coder0', 'coder0-0']);
+    expect(peers.find((p) => p.name === 'coder0')).toMatchObject({ className: 'implementer' });
+    expect(peers.find((p) => p.name === 'coder0-0')).toMatchObject({
+      className: 'implementer',
+      host: '10.1.10.37',
+      ssh: 'jwoltje@10.1.10.37',
+    });
+    // local agents have no host/ssh
+    expect(peers.find((p) => p.name === 'orchestrator')!.host).toBeUndefined();
+  });
+
+  it('parses an optional per-agent socket', () => {
+    const peers = parseRosterAgents(
+      ['agents:', '  - name: a', '    class: worker', '    socket: mosaic-factory'].join('\n'),
+    );
+    expect(peers[0]).toMatchObject({ name: 'a', socket: 'mosaic-factory' });
+  });
+
+  it('stops at the next top-level key', () => {
+    const peers = parseRosterAgents(
+      ['agents:', '  - name: a', '    class: worker', 'defaults:', '  working_directory: ~'].join(
+        '\n',
+      ),
+    );
+    expect(peers.map((p) => p.name)).toEqual(['a']);
+  });
+});
+
+describe('renderPeerReach — same-host vs cross-host', () => {
+  const send = '/home/u/.config/mosaic/tools/tmux/agent-send.sh';
+
+  it('renders the short form for a same-host peer', () => {
+    const peer: CommsPeer = { name: 'enhancer', className: 'enhancer' };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(`${send} -s enhancer -m "…"`);
+  });
+
+  it('renders the -H form for a cross-host peer using ssh', () => {
+    const peer: CommsPeer = {
+      name: 'coder0-0',
+      className: 'implementer',
+      host: '10.1.10.37',
+      ssh: 'jwoltje@10.1.10.37',
+    };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(
+      `${send} -H jwoltje@10.1.10.37 -s coder0-0 -m "…"`,
+    );
+  });
+
+  it('falls back to host when a cross-host peer has no ssh', () => {
+    const peer: CommsPeer = { name: 'x', className: 'worker', host: '10.0.0.9' };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(`${send} -H 10.0.0.9 -s x -m "…"`);
+  });
+
+  it('treats a peer whose host equals the fleet host as same-host', () => {
+    const peer: CommsPeer = { name: 'y', className: 'worker', host: 'w-jarvis' };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(`${send} -s y -m "…"`);
+  });
+
+  it('emits NO -L for an unset/default socket', () => {
+    const peer: CommsPeer = { name: 'lead', className: 'orchestrator' };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(`${send} -s lead -m "…"`);
+  });
+
+  it('emits -L <socket> for a named socket', () => {
+    const peer: CommsPeer = { name: 'coder0', className: 'implementer', socket: 'mosaic-factory' };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(
+      `${send} -L mosaic-factory -s coder0 -m "…"`,
+    );
+  });
+
+  it('combines -L (named socket) and -H (cross-host) in order', () => {
+    const peer: CommsPeer = {
+      name: 'coder0-0',
+      className: 'implementer',
+      host: '10.1.10.37',
+      ssh: 'jwoltje@10.1.10.37',
+      socket: 'mosaic-factory',
+    };
+    expect(renderPeerReach(peer, 'w-jarvis', send)).toBe(
+      `${send} -L mosaic-factory -H jwoltje@10.1.10.37 -s coder0-0 -m "…"`,
+    );
+  });
+});
+
+describe('buildFleetCommsBlock', () => {
+  const send = '/h/.config/mosaic/tools/tmux/agent-send.sh';
+  const agents = parseRosterAgents(ROSTER);
+
+  it('excludes self, lists peers, flags the orchestrator, and emits both address forms', () => {
+    const block = buildFleetCommsBlock({
+      selfName: 'enhancer',
+      agents,
+      fleetHost: 'w-jarvis',
+      agentSendPath: send,
+    });
+    expect(block).toContain('# Fleet Comms');
+    expect(block).toContain('You are **enhancer**');
+    // criterion 1: agent's own [host:session] identity
+    expect(block).toContain('`[w-jarvis:enhancer]`');
+    // self excluded
+    expect(block).not.toMatch(/\|\s*enhancer\s*\|/);
+    // peers present
+    expect(block).toContain('| orchestrator |');
+    expect(block).toContain('point of contact');
+    // same-host peer short form
+    expect(block).toContain(`${send} -s coder0 -m "…"`);
+    // cross-host peer -H form + host annotation
+    expect(block).toContain(`${send} -H jwoltje@10.1.10.37 -s coder0-0 -m "…"`);
+    expect(block).toContain('host `10.1.10.37`');
+    // conventions
+    expect(block).toContain('FLIP the preamble');
+    expect(block).toContain('ACCEPTED');
+  });
+
+  it('returns empty when the agent has no peers', () => {
+    expect(
+      buildFleetCommsBlock({
+        selfName: 'solo',
+        agents: [{ name: 'solo', className: 'orchestrator' }],
+        fleetHost: 'h',
+        agentSendPath: send,
+      }),
+    ).toBe('');
+  });
+});
+
+describe('readFleetCommsBlock — situational (the context a spawned agent gets)', () => {
+  let home: string;
+  beforeEach(() => {
+    home = mkdtempSync(join(tmpdir(), 'mosaic-comms-'));
+    mkdirSync(join(home, 'fleet'), { recursive: true });
+    writeFileSync(join(home, 'fleet', 'roster.yaml'), ROSTER);
+  });
+  afterEach(() => rmSync(home, { recursive: true, force: true }));
+
+  it('builds the cheat-sheet with correct peer addresses for a fleet member', () => {
+    const block = readFleetCommsBlock(home, 'orchestrator', 'w-jarvis');
+    expect(block).toContain('# Fleet Comms');
+    expect(block).toContain('| enhancer |');
+    expect(block).toContain(`${join(home, 'tools', 'tmux', 'agent-send.sh')} -s coder0 -m "…"`);
+    expect(block).toContain('-H jwoltje@10.1.10.37 -s coder0-0');
+    expect(block).not.toMatch(/\|\s*orchestrator\s*\|/); // self excluded
+  });
+
+  it('returns empty when MOSAIC_AGENT_NAME is unset, no roster, or agent not a member', () => {
+    expect(readFleetCommsBlock(home, undefined, 'w-jarvis')).toBe('');
+    expect(readFleetCommsBlock(home, 'stranger', 'w-jarvis')).toBe('');
+    expect(readFleetCommsBlock(mkdtempSync(join(tmpdir(), 'noroster-')), 'orchestrator')).toBe('');
+  });
+});

packages/mosaic/src/fleet/comms-onboarding.ts

@@ -0,0 +1,183 @@
+/**
+ * Fleet onboarding-injection (#620).
+ *
+ * Fleet agents are born not knowing how to reach their peers — the root cause of
+ * a spawned agent's failed first send. When an agent boots via `mosaic yolo
+ * <runtime>` (→ composeContract → system prompt), we append a comms cheat-sheet
+ * + peer roster so it can talk to the orchestrator and other agents immediately.
+ *
+ * Cross-host aware: a peer may carry `host`/`ssh` (a deliberate pre-federation
+ * stopgap — manual cross-host listing; federation/W1 auto-discovers later), so a
+ * w-jarvis agent is born knowing the exact `-H` command to reach a dragon-lin
+ * peer. Same-host peers render the short form.
+ *
+ * Standalone (no fleet.ts import) to keep launch.ts's prompt path free of the
+ * heavy fleet command module. The roster is parsed leniently — the cheat-sheet
+ * is best-effort onboarding, never a hard dependency.
+ */
+
+import { existsSync, readFileSync } from 'node:fs';
+import { homedir, hostname } from 'node:os';
+import { join } from 'node:path';
+
+export interface CommsPeer {
+  name: string;
+  /** Roster `class` (orchestrator | enhancer | implementer | worker | …). */
+  className: string;
+  /** Host the peer runs on; absent ⇒ the fleet host (same host). */
+  host?: string;
+  /** SSH target (user@host) for a cross-host peer; renders the `-H` form. */
+  ssh?: string;
+  /** tmux socket the peer's session lives on; absent ⇒ default socket (no `-L`). */
+  socket?: string;
+}
+
+/**
+ * Lenient parse of a fleet `roster.yaml` for agent name/class/host/ssh. Avoids a
+ * dependency on the full fleet roster parser; the format is `- name:` list items
+ * with `class:`/`host:`/`ssh:` siblings under `agents:`.
+ */
+export function parseRosterAgents(yamlText: string): CommsPeer[] {
+  const peers: CommsPeer[] = [];
+  let current: CommsPeer | null = null;
+  let inAgents = false;
+  const scalar = (line: string, key: string): string | null => {
+    const m = line.match(new RegExp(`^\\s*${key}:\\s*["']?([^"'#]+?)["']?\\s*$`));
+    return m ? (m[1] as string).trim() : null;
+  };
+  for (const rawLine of yamlText.split('\n')) {
+    const line = rawLine.replace(/\s+$/, '');
+    if (/^agents:\s*$/.test(line)) {
+      inAgents = true;
+      continue;
+    }
+    if (!inAgents) continue;
+    // A new top-level key (no leading space) ends the agents block.
+    if (/^\S/.test(line)) break;
+
+    const nameMatch = line.match(/^\s*-\s*name:\s*["']?([A-Za-z0-9._-]+)["']?\s*$/);
+    if (nameMatch) {
+      if (current) peers.push(current);
+      current = { name: nameMatch[1] as string, className: 'worker' };
+      continue;
+    }
+    if (!current) continue;
+    const cls = scalar(line, 'class');
+    if (cls) current.className = cls;
+    const host = scalar(line, 'host');
+    if (host) current.host = host;
+    const ssh = scalar(line, 'ssh');
+    if (ssh) current.ssh = ssh;
+    const socket = scalar(line, 'socket');
+    if (socket) current.socket = socket;
+  }
+  if (current) peers.push(current);
+  return peers;
+}
+
+export interface FleetCommsOptions {
+  /** This agent's name (it is excluded from its own peer list). */
+  selfName: string;
+  /** All roster agents (including self; filtered out internally). */
+  agents: CommsPeer[];
+  /** Host the fleet runs on (short hostname) — the same-host baseline. */
+  fleetHost: string;
+  /** Absolute path to agent-send.sh in this install. */
+  agentSendPath: string;
+}
+
+/** Is this peer on a different host than the fleet baseline? */
+function isRemote(peer: CommsPeer, fleetHost: string): boolean {
+  return peer.host !== undefined && peer.host !== fleetHost;
+}
+
+/**
+ * Render the exact agent-send command to reach a peer (session = agent name).
+ * Data-driven per peer: a named `socket` → `-L <socket>`; an unset socket → the
+ * default tmux socket (no `-L`). A cross-host peer adds `-H <ssh|host>`.
+ */
+export function renderPeerReach(peer: CommsPeer, fleetHost: string, agentSendPath: string): string {
+  const parts = [agentSendPath];
+  if (peer.socket) parts.push('-L', peer.socket); // unset ⇒ default socket, no -L
+  if (isRemote(peer, fleetHost)) parts.push('-H', peer.ssh ?? (peer.host as string));
+  parts.push('-s', peer.name, '-m', '"…"');
+  return parts.join(' ');
+}
+
+/**
+ * Build the `# Fleet Comms` onboarding block (pure markdown). Returns '' when
+ * the agent has no peers (a single-agent roster has no one to talk to).
+ */
+export function buildFleetCommsBlock(opts: FleetCommsOptions): string {
+  const peers = opts.agents.filter((a) => a.name !== opts.selfName);
+  if (peers.length === 0) return '';
+
+  const orchestrator = peers.find((p) => p.className === 'orchestrator');
+  const rows = peers
+    .map((p) => {
+      const where = isRemote(p, opts.fleetHost)
+        ? `${p.className} · host \`${p.host}\``
+        : p.className;
+      const role = p.className === 'orchestrator' ? `${where} ← point of contact` : where;
+      return `| ${p.name} | ${role} | \`${renderPeerReach(p, opts.fleetHost, opts.agentSendPath)}\` |`;
+    })
+    .join('\n');
+
+  const orchLine = orchestrator
+    ? `Your point of contact is **${orchestrator.name}** (the orchestrator) — route questions, ` +
+      `status, and decisions there.`
+    : `This fleet has no orchestrator in its roster; coordinate with your peers directly.`;
+
+  return `# Fleet Comms — reach your peers
+
+You are **${opts.selfName}** in this fleet. Your comms identity is \`[${opts.fleetHost}:${opts.selfName}]\` —
+that is the \`<src>\` other agents see and reply to. Reach other agents (durable tmux sessions) with the
+Mosaic comms tool at \`${opts.agentSendPath}\`. The **Reach** column below is the exact command per peer:
+same-host peers use the short form (no \`-H\`); cross-host peers include \`-H <user@host>\`.
+
+## Peers
+
+| Agent | Role | Reach (session = agent name) |
+| ----- | ---- | ---------------------------- |
+${rows}
+
+${orchLine}
+
+## Conventions
+
+- Every message carries a self-identifying preamble \`[<src_host>:<src_session> -> <dst_host>:<dst_session>]\` — \`agent-send.sh\` adds it automatically.
+- **To reply, FLIP the preamble:** address your reply to the sender's \`src\` (their host:session becomes your \`-s\`/\`-H\`).
+- \`agent-send.sh\` (a.k.a. \`agent send --verify\`) confirms the message was **ACCEPTED** at the destination prompt — not merely injected. Prefer it for anything that matters.`;
+}
+
+/**
+ * Read the fleet roster from `mosaicHome` and build the comms block for
+ * `selfName`. Returns '' when there is no roster, the agent is not in it, or
+ * there are no peers — onboarding is best-effort and never throws.
+ */
+export function readFleetCommsBlock(
+  mosaicHome: string,
+  selfName: string | undefined,
+  fleetHost: string = hostname().split('.')[0] || 'localhost',
+): string {
+  if (!selfName) return '';
+  const rosterPath = join(mosaicHome, 'fleet', 'roster.yaml');
+  if (!existsSync(rosterPath)) return '';
+  let text: string;
+  try {
+    text = readFileSync(rosterPath, 'utf-8');
+  } catch {
+    return '';
+  }
+  const agents = parseRosterAgents(text);
+  if (!agents.some((a) => a.name === selfName)) return ''; // not a member of this fleet
+  return buildFleetCommsBlock({
+    selfName,
+    agents,
+    fleetHost,
+    agentSendPath: join(mosaicHome, 'tools', 'tmux', 'agent-send.sh'),
+  });
+}
+
+/** Default mosaic home (mirrors launch.ts), for callers that don't pass one. */
+export const DEFAULT_MOSAIC_HOME_FOR_COMMS = join(homedir(), '.config', 'mosaic');