Files
gbrain/recipes/agent-voice/code/gbrain-client.mjs
T
e9fa51d46e v0.40.0.0 feat: agent-voice (Mars + Venus) + copy-into-host-repo skillpack paradigm (#1128)
* feat: agent-voice reference skillpack (Mars + Venus) + copy-into-host-repo install paradigm

Ships a new skillpack paradigm: gbrain holds the REFERENCE content;
`gbrain integrations install agent-voice --target <repo>` COPIES it into
the operator's host agent repo where it becomes user-owned and mutable.
Future refresh is diff-and-propose against per-file SHA-256 hashes from
.gbrain-source.json, not blind overwrite.

What ships:
- recipes/agent-voice.md entrypoint + recipes/agent-voice/ bundle
- Two voice personas (Mars dual-mode SOLO/DEMO, Venus executive assistant)
  with PII / private-agent-name / hardcoded-path scrubbed out
- WebRTC-first browser client (call.html) with ?test=1 gated instrumentation
  and Web Audio API tee -> MediaRecorder capture for E2E roundtrip testing
- Read-only tool router (D14-A allow-list: search, query, get_page,
  list_pages, find_experts, get_recent_salience, get_recent_transcripts,
  read_article). Write ops permanently denylisted; opt-in via local override
- Persona-aware prompt builder with identity-first composition + Unicode
  sanitization for OpenAI Realtime API safety
- Upstream-error classifier (HTTP 429/500/503 -> soft-fail, plumbing -> hard)
- Three SKILL.md skills (voice-persona-mars, voice-persona-venus,
  voice-post-call) with routing-eval.jsonl fixtures
- 99 host-side tests (vitest-compatible, runs in bun) covering registry,
  prompt-shape privacy guards, tool allow-list, upstream classifier
- install/manifest.json + refresh-algorithm.md + post-install-hint.md

Privacy infrastructure:
- scripts/check-no-pii-in-agent-voice.sh wired into bun run verify
  Shape regex (phone/email/SSN/JWT/bearer/credit-card) + path patterns +
  $AGENT_VOICE_PII_BLOCKLIST env-driven name blocklist
- scripts/import-from-upstream.sh + scripts/upstream-scrub-table.txt
  Deterministic refresh from upstream voice-agent source. Placeholder-
  driven (envsubst-expanded at run time) so no private names land in
  checked-in files
- recipes/agent-voice/code/lib/personas/private-name-blocklist.json
  Single source of truth for the regex contract (shape categories +
  path patterns + env-var contract for operator-specific names)

src/ surface:
- src/commands/integrations.ts gains `install <recipe-id>` subcommand
  with install_kind: 'local-managed' | 'copy-into-host-repo' discriminator.
  Path-traversal hardening (rejects '..', absolute paths, symlink escapes).
  Refuses target == gbrain itself, missing .git, existing files (without
  --overwrite). Writes .gbrain-source.json with per-file SHA-256. Appends
  resolver rows to host repo's RESOLVER.md or AGENTS.md.
- test/integrations-install.test.ts: 11 cases (happy path, manifest shape,
  no upstream_repo field per D11-A, resolver appending, file modes,
  refusal cases, dry-run)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* chore: bump version and changelog (v0.36.0.0)

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* fix(privacy): scrub literal private agent names from prompt-shape tests + guard script

The prompt-shape tests carried regex patterns naming the literal banned terms
(Garry/Steph/Garrison/Solomon/Herbert/Wintermute) inline. CLAUDE.md's
"never use Wintermute in any public artifact" applies to test source files
too. Master's check-privacy.sh correctly caught this.

Replaced with env-driven check that reads AGENT_VOICE_PII_BLOCKLIST (the
single source of truth from private-name-blocklist.json). Same enforcement
guarantee via the env var, zero literal names in shipped source.

Also scrubbed the literal /data/.openclaw/ from the guard script's comment
and the literal 'tell_wintermute' from the venus write-tools test.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* feat: ship all v0.36.0.0 deferred items in this PR (E2E + evals + pipeline + refresh + multilingual + twilio deprecation)

Closes the "deferred to follow-up" section of the v0.36.0.0 CHANGELOG.

E2E tests + harness (env-gated):
- tests/e2e/voice-roundtrip.test.mjs — spawns server, drives puppeteer + fake-audio, three-tier assertions (CONNECTION hard, NON-SILENT hard, SEMANTIC soft via Whisper + LLM judge). Upstream errors (429/500/503, WS 1011/1013) soft-fail via lib/upstream-classifier.mjs.
- tests/e2e/voice-full-flow.test.mjs — wraps openclaw doing the install, then runs the roundtrip. Friction-discovery flavor, NOT a ship gate.
- tests/e2e/lib/browser-audio.mjs — puppeteer + fake-audio harness; reads window._gbrainTest namespace; PCM RMS-variance helper.
- tests/e2e/lib/whisper-judge.mjs — Whisper transcription + LLM-judge for SEMANTIC tier.
- tests/e2e/audio-fixtures/utterance-{add,joke,brain-query}.wav — 16kHz mono WAV via `say` + ffmpeg, committed for reproducibility.
- test/fixtures/claw-test-scenarios/voice-agent-install/{BRIEF.md, scenario.json, expected.json} — labeled BENCHMARK_FRICTION, blocks_ship=false.

LLM-judge persona evals + synthetic canonical baselines:
- tests/evals/judge.mjs — gateway-routed 3-model (Claude + GPT + Gemini) harness with 4-strategy JSON repair + 2/3-quorum aggregation (per the v0.27.x cross-modal pattern). Pass criterion: every axis mean ≥7 AND no model <5.
- tests/evals/fixtures/{mars-solo,mars-demo,venus,persona-routing,mars-multilingual}.jsonl — 5 fixture sets covering all axes.
- tests/evals/{mars-eval,venus-eval,mars-multilingual-eval,persona-routing-eval}.mjs — per-axis drivers.
- tests/evals/baseline-runs/canonical/*.json — agent-authored synthetic exemplars (PII-impossible by construction; demonstrate expected pass shape; never overwrite with live model output).
- tests/evals/baseline-runs/.gitignore — live receipts excluded.

DIY pipeline (Option B):
- code/pipeline.mjs — streaming STT (Deepgram nova-2) + LLM (Claude Sonnet 4.6 streaming SSE with sentence-boundary TTS dispatch) + TTS (Cartesia primary, OpenAI TTS fallback). 20-turn history cap, exponential-backoff reconnects, 25s keepalives, VAD presets (quiet/normal/noisy/very_noisy), barge-in via STT speechStart → LLM interrupt. Modular adapters for swapping providers.

--refresh mode (D3-A diff-and-propose):
- src/commands/integrations.ts: refreshRecipeIntoHostRepo() + classifyForRefresh() implementing the five states from refresh-algorithm.md (unchanged-identical, unchanged-stale, locally-modified, source-deleted, host-deleted, new-in-manifest). Transaction journal at .gbrain-source.refresh.log. Default policy: preserve operator's local edits (keep-mine); --auto take-theirs to overwrite; --dry-run for preview.
- test/integrations-install.test.ts: 7 new test cases pinning each classification state + default-preserve behavior + take-theirs overwrite + transaction journal + refusal on uninstalled target.

Mars multilingual restore:
- code/lib/personas/mars.mjs: explicit cross-lingual rule (Mandarin, Spanish, French, Japanese, Korean default to English but follow the speaker). Voice (Orus) supports the languages natively.
- tests/unit/mars-prompt-shape.test.mjs: assertion flipped from "MUST NOT claim multilingual" to "declares cross-lingual capability with English bias."
- tests/evals/fixtures/mars-multilingual.jsonl: 5 fixtures across Mandarin/Spanish/Japanese/French + explicit switch-back, pinned by mars-multilingual-eval.mjs.

Twilio recipe deprecation:
- recipes/twilio-voice-brain.md: deprecation banner pointing at agent-voice.md. Frontmatter version bumped to 0.8.2. Will be removed in v0.37.

Verify: bun run verify clean, 6736+ unit tests pass, 18/18 install+refresh tests pass, 96/98 host-side persona/tool/classifier tests pass (2 skipped env-gated).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* docs: update CHANGELOG — all v0.36.0.0 deferred items now shipped in this PR

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* chore: rebump version v0.36.0.0 → v0.37.0.0

Captures the wave-1 + wave-2 scope at the v0.37 slot. The bump reflects
the size of what this PR ships: copy-into-host-repo install paradigm
(new install_kind discriminator + new install/refresh subcommand) +
Mars/Venus voice agent reference + 5,500+ LOC of vendored scrubbed
code + 4 LLM-judge eval suites + 2 env-gated E2E test suites + DIY
Option B pipeline + 18-case install subcommand test coverage. A minor
bump felt too small.

Side fix: privacy guard caught two stale literal "wintermute" and
"/data/.openclaw/" references in wave-2 files
(voice-full-flow.test.mjs comment, expected.json blocklist payload).
Both replaced with env-driven references to $AGENT_VOICE_PII_BLOCKLIST
matching the D15-A pattern from the original review.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

* chore: rebump v0.37.0.0 → v0.40.0.0

Jumps past the v0.37/v0.38/v0.39 slots master might claim in subsequent
PRs. The wave's scope (copy-into-host-repo skillpack paradigm + agent-voice
+ install/refresh + LLM-judge evals + DIY pipeline + Mars multilingual)
justifies a larger version arithmetic step.

Files bumped:
- VERSION 0.37.0.0 → 0.40.0.0
- package.json 0.37.0.0 → 0.40.0.0
- CHANGELOG.md header + "To take advantage of v0.40.0.0" block
- recipes/twilio-voice-brain.md deprecation banner (now "removed in v0.41")
- recipes/agent-voice/tests/evals/mars-multilingual-eval.mjs comment

Left alone: master's pre-existing "v0.37+" roadmap labels in src/core/calibration/*,
src/core/cycle/*, DESIGN.md, CLAUDE.md, etc. Those are master's author-intent
references to "the next planned release" relative to master's frame at the time —
rewriting them just to keep numbering consistent would overreach.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-22 21:54:13 -07:00

183 lines
5.5 KiB
JavaScript

/**
* gbrain-client.mjs — minimal stdio MCP client for gbrain.
*
* Spawns `gbrain serve` as a long-lived child process and communicates via
* JSON-RPC over stdio. The client is intentionally minimal: open the
* connection once, send `initialize`, then forward tool calls. No retries,
* no batching, no streaming results (most voice tools return small
* payloads; large ones can be paginated by the caller).
*
* Production hardening this file does NOT do:
* - reconnect on child crash (caller restarts; voice sessions are short)
* - request timeouts (caller handles via its own AbortSignal)
* - concurrent request multiplexing (one in-flight call at a time)
*
* That's deliberate: this is reference code optimized for clarity and the
* voice-agent use case, not a production-grade MCP client. The operator
* can replace it with a richer implementation (e.g., the
* @modelcontextprotocol/sdk Client) without changing tools.mjs.
*
* Configuration: `$GBRAIN_BIN` (default: `gbrain` on PATH) is the binary to
* spawn. `$GBRAIN_BRAIN_ID` and `$GBRAIN_SOURCE` route to a specific brain
* + source if set (see the gbrain docs/architecture/brains-and-sources.md).
*/
import { spawn } from 'node:child_process';
const GBRAIN_BIN = process.env.GBRAIN_BIN || 'gbrain';
let _child;
let _nextId = 1;
const _pending = new Map();
let _initialized = false;
let _buffer = '';
function ensureChild() {
if (_child && !_child.killed) return _child;
const args = ['serve'];
if (process.env.GBRAIN_BRAIN_ID) args.push('--brain', process.env.GBRAIN_BRAIN_ID);
if (process.env.GBRAIN_SOURCE) args.push('--source', process.env.GBRAIN_SOURCE);
_child = spawn(GBRAIN_BIN, args, {
stdio: ['pipe', 'pipe', 'pipe'],
env: { ...process.env, MCP_STDIO: '1' },
});
_child.stdout.setEncoding('utf8');
_child.stdout.on('data', (chunk) => {
_buffer += chunk;
// JSON-RPC over stdio: messages separated by newlines.
let nlIdx;
while ((nlIdx = _buffer.indexOf('\n')) !== -1) {
const line = _buffer.slice(0, nlIdx).trim();
_buffer = _buffer.slice(nlIdx + 1);
if (!line) continue;
try {
const msg = JSON.parse(line);
if (msg.id !== undefined && _pending.has(msg.id)) {
const { resolve, reject } = _pending.get(msg.id);
_pending.delete(msg.id);
if (msg.error) {
reject(new Error(msg.error.message || 'gbrain MCP error'));
} else {
resolve(msg.result);
}
}
// notifications/log etc. — ignore for the v0 voice use case.
} catch (err) {
// Malformed line; ignore. gbrain prints structured JSON-RPC only,
// but a renegade stderr-bleed could surface here.
}
}
});
_child.stderr.setEncoding('utf8');
_child.stderr.on('data', (chunk) => {
// Surface gbrain stderr to our stderr for debugging. Don't crash on it.
process.stderr.write(`[gbrain] ${chunk}`);
});
_child.on('exit', (code, signal) => {
const reason = signal ? `signal ${signal}` : `code ${code}`;
for (const [, p] of _pending) {
p.reject(new Error(`gbrain child exited (${reason}) with ${_pending.size} pending`));
}
_pending.clear();
_initialized = false;
_child = null;
});
return _child;
}
function rpc(method, params) {
const id = _nextId++;
const child = ensureChild();
return new Promise((resolve, reject) => {
_pending.set(id, { resolve, reject });
const msg = JSON.stringify({ jsonrpc: '2.0', id, method, params }) + '\n';
try {
child.stdin.write(msg);
} catch (err) {
_pending.delete(id);
reject(err);
}
});
}
async function initIfNeeded() {
if (_initialized) return;
await rpc('initialize', {
protocolVersion: '2024-11-05',
capabilities: {},
clientInfo: { name: 'agent-voice', version: '0.1.0' },
});
_initialized = true;
}
/**
* Call a single gbrain operation by name with params. Returns the operation
* result (whatever shape gbrain returns; usually a JSON-serializable object).
*
* @throws {Error} on transport failure or operation error.
*/
export async function callGbrainOp(opName, params) {
await initIfNeeded();
// gbrain exposes operations as MCP "tools." The standard call is
// tools/call with {name, arguments}.
const result = await rpc('tools/call', { name: opName, arguments: params || {} });
// gbrain returns {content: [{type:'text', text:'...JSON...'}]}; parse if needed.
if (result?.content?.[0]?.type === 'text') {
const text = result.content[0].text;
try {
return JSON.parse(text);
} catch {
return text;
}
}
return result;
}
/**
* Test hook: stub the underlying RPC. Tests use this instead of spawning a
* real gbrain child. Pass `null` to restore the real spawn-and-rpc path.
*/
export function __setRpcForTests(fn) {
if (fn === null) {
_testRpc = null;
return;
}
_testRpc = fn;
}
let _testRpc = null;
// Re-export rpc through the test hook for callGbrainOp.
const _origRpc = rpc;
async function dispatchRpc(method, params) {
if (_testRpc) return _testRpc(method, params);
return _origRpc(method, params);
}
// Override the helper used by callGbrainOp.
export async function _testableCallGbrainOp(opName, params) {
if (_testRpc) {
return _testRpc('tools/call', { name: opName, arguments: params || {} });
}
return callGbrainOp(opName, params);
}
/**
* Shutdown helper. Tests call this between cases; production never has to.
*/
export function shutdown() {
if (_child) {
_child.kill();
_child = null;
}
_pending.clear();
_initialized = false;
_buffer = '';
}