Files
gbrain/scripts/check-no-pii-in-agent-voice.sh
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

115 lines
4.6 KiB
Bash
Executable File

#!/usr/bin/env bash
# CI guard: scan agent-voice shipped surface for privacy leaks.
#
# Catches three classes of leak:
# 1. SHAPE regex: phone, email, SSN, JWT, bearer token, Luhn-valid credit card.
# Mirrors the patterns in src/core/eval-capture-scrub.ts.
# 2. PATH patterns: hardcoded private filesystem prefixes (e.g. private agent home dirs).
# 3. OPERATOR blocklist: pipe-separated word list from $AGENT_VOICE_PII_BLOCKLIST.
# Operator/CI sets this to ban specific private names without committing them
# to a public file. See recipes/agent-voice/code/lib/personas/private-name-blocklist.json
# for the contract.
#
# Exit 0 on no matches; exit 1 on any match. Wired into `bun run verify`.
#
# Test fixtures under recipes/agent-voice/tests/fixtures/scrub-{dirty,clean}.txt are
# the deliberate dirty inputs used by test/check-no-pii.test.ts. They contain the
# token FORBIDDEN_PLACEHOLDER_NAME_1 (NOT real names) and the test sets
# AGENT_VOICE_PII_BLOCKLIST=FORBIDDEN_PLACEHOLDER_NAME_1 to verify the mechanism.
set -euo pipefail
ROOT="$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
cd "$ROOT"
SCAN_PATHS=(
"recipes/agent-voice"
"recipes/agent-voice.md"
"scripts/import-from-upstream.sh"
"scripts/upstream-scrub-table.txt"
)
# Exception list: files the guard deliberately skips. These are test fixtures that
# carry placeholder tokens to exercise the guard's own behavior.
EXCEPTION_PATHS=(
"recipes/agent-voice/tests/fixtures/scrub-dirty.txt"
"recipes/agent-voice/tests/fixtures/scrub-clean.txt"
)
# Build a single grep --exclude argument list from EXCEPTION_PATHS.
EXCLUDES=()
for p in "${EXCEPTION_PATHS[@]}"; do
EXCLUDES+=("--exclude=$(basename "$p")")
done
# SHAPE regex — keep in sync with src/core/eval-capture-scrub.ts. Each pattern
# is wrapped in (?: ... ) so we can OR them into one combined regex.
# Note: we use grep -E (POSIX ERE), which doesn't support lookbehind. The
# patterns here are slightly looser than the JS RegExps in eval-capture-scrub.ts
# because lookbehind is unavailable; that loosening errs on the side of catching
# more, which is the safer failure mode for a privacy guard.
PHONE='(\+[0-9]{1,3}[ .-]?)?(\([0-9]{3}\)[ ]?|[0-9]{3}[ .-])[0-9]{3}[ .-]?[0-9]{4}'
EMAIL='[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}'
SSN='[0-9]{3}-[0-9]{2}-[0-9]{4}'
JWT='\beyJ[A-Za-z0-9_-]{8,}\.[A-Za-z0-9_-]{8,}\.[A-Za-z0-9_-]{8,}\b'
BEARER='([Bb]earer)[ ]+[A-Za-z0-9._~+/-]{10,}=*'
# Credit card: 13-19 digits with optional space/dash separators. Luhn check NOT
# implemented in this guard (the grep is shape-only). False positives expected
# on long numeric sequences; treat as "investigate, this is privacy-shaped."
CC='([0-9][ -]?){12,18}[0-9]'
# PATH patterns — hardcoded private filesystem prefixes. Add new ones here as
# new private deployment patterns emerge.
PATH_PATTERNS='(/data/\.openclaw/|/private/[a-z0-9_-]+/workspace/)'
# Combine SHAPE + PATH into one ERE.
COMBINED_REGEX="($PHONE)|($EMAIL)|($SSN)|($JWT)|($BEARER)|($CC)|($PATH_PATTERNS)"
# Track any failure across the script.
FAILED=0
scan_pattern() {
local label="$1"
local pattern="$2"
local hits
# -r recursive, -E extended regex, -n line numbers, -I skip binary files.
# ${SCAN_PATHS[@]} is intentionally unquoted (we want word-splitting for
# path-list expansion). Existing-path filter via test -e.
local existing=()
for p in "${SCAN_PATHS[@]}"; do
if [ -e "$p" ]; then existing+=("$p"); fi
done
if [ ${#existing[@]} -eq 0 ]; then
return 0
fi
if hits=$(grep -rEnI "${EXCLUDES[@]}" "$pattern" "${existing[@]}" 2>/dev/null); then
echo
echo "ERROR: $label leak detected in agent-voice surface:"
echo "$hits"
FAILED=1
fi
}
scan_pattern "shape/path PII" "$COMBINED_REGEX"
# Operator blocklist via env var. If set, run a second scan.
if [ -n "${AGENT_VOICE_PII_BLOCKLIST:-}" ]; then
# The env var is pipe-separated. Build a case-insensitive grep -E pattern.
# \b word boundaries to avoid matching substrings ("garrison" matching inside
# a longer benign word).
OPERATOR_PATTERN="\\b(${AGENT_VOICE_PII_BLOCKLIST})\\b"
scan_pattern "operator blocklist match" "$OPERATOR_PATTERN"
else
echo "WARN: AGENT_VOICE_PII_BLOCKLIST env var is unset; running shape-only PII scan." >&2
echo " For full enforcement, set the env var to a pipe-separated word list." >&2
echo " See recipes/agent-voice/code/lib/personas/private-name-blocklist.json." >&2
fi
if [ "$FAILED" -ne 0 ]; then
echo
echo "FAIL: agent-voice privacy guard found leaks. Scrub before commit."
exit 1
fi
echo "OK: agent-voice surface is clean (shape + path + operator blocklist)"