BoB v0.98.0

BoB v0.98.0 — the memory-hardening release

Added

• Single-writer memory write fence (OS-level, fail-closed). Every memory writer must hold a family-scoped lock (one per vector-store service + collection family, spanning all vector dims). Memory enabled with no armed fence refuses to boot; a second writer degrades to read-only with honest health (HTTP 423 memory_write_locked) instead of corrupting the store; permission problems report as permission problems.
• Experimental zero-Docker vector store (opt-in). The zvec embedded store runs as a supervised child process owned by core (crash → clean error, next call restarts it). Opt in via config/memory_stores.toml + pip install zvec==0.5.1; Docker Qdrant remains the default.
• BobLKS: a local corpus instance — ingest() / retrieve() over your own documents through the same fence, fingerprint, and provider seams as agent memory.
• Asymmetric embedder seam — typed embed_query / embed_doc with per-slot instruction templates (part of the embed fingerprint), batching, and a first-write dimension probe (writes fail closed on mismatch, reads fail open).
• Retrieval eval harness with an honest ceiling (bobclaw-core/evals/retrieval/): author-blind paraphrase set, Wilson CIs, paired outputs. Measured recall@10: qwen3-embedding-4b 65%, 0.6b 40%. Earlier 95%/88% figures were lexical-leakage-inflated and are relabeled as an embedding smoke.
• Recall backfill past dangling vectors — bounded overfetch/paging to reach top_k valid facts instead of silently truncating.
• CI + release-surface baseline runner — .github/workflows/tests.yml (Windows + Linux from the pinned locks) and ./run_baseline_tests.ps1, which exits non-zero if ANY suite fails.
• UPGRADE.md — the supported tag-to-tag upgrade runbook.

Changed

• Default embedder: granite-embedding-311m → qwen3-embedding-4b (768 → 2560 dims, last-token pooling), decided on the paraphrase eval above. granite remains the documented CPU-light alternative (not yet measured on the new eval set — an open follow-up). Existing memory stores must re-index — see UPGRADE.md.
• Compose project naming is pinned and fixed container names are removed, so two installs coexist on one host (distinct COMPOSE_PROJECT_NAME + the new BOBCLAW_*_PORT datastore-port overrides). Existing installs: see UPGRADE.md before your next compose up.
• L1 fact dedup identity excludes volatile turn metadata (extractor v1→v2), so a repeated fact no longer multiplies.

Fixed

• Qdrant container healthcheck was always failing (image ships no wget) — now probes /healthz over bash /dev/tcp.
• MEMORY_SLOTS_FILE was documented but ignored by bootstrap — a second install could not point its embedder/extractor slots away from the shipped defaults. Now honored (with a regression test).
• start-local.ps1 probed fixed ports for its idempotence check and could mistake a coexisting install's gateway for its own — now probes the ports the install is configured for.
• Memory integration smokes isolate to throwaway collections with a fail-closed write guard.

Known limitations (honest)

• zvec approximate search trails exact search on recall (21/40 vs 26/40 through the full production path); a rescoring tier is planned for v0.99 — Qdrant stays the recommended default.
• The document parser applies soft limits only (hard cap filed for v0.99).
• pip-audit reports UNAVAILABLE (never silently "clean") when not installed.

Full notes in CHANGELOG.md. Release gate: clean-clone, shipped-install-path baseline — core 2846 passed / 0 failed / 6 skipped, gateway / pipeline / KMM / pip-check green — plus a live production-shaped stand-up (login, chat, fan-out, memory write + recall, restart survival) on a second isolated install.

BoB v0.97.0

BoB v0.97.0

Added

• Chinese localization (Simplified + Traditional). The desktop app UI in en / zh-Hans (简) / zh-Hant (繁) with a restart-free header language toggle, plus a backend per-turn locale directive (English path byte-identical).
• Codex planner honors an explicitly-pinned model. When a specific model is pinned (switch_model / state.model_override), a gpt-profile face runs that chosen GPT model natively under a ChatGPT login instead of only the profile default — no LiteLLM proxy. (A GUI model-picker control is a follow-up.)
• Faces know they're running inside BoB. An opt-in spawn-identity card names the platform, the face (name / role), and the backend serving it. Code default off (byte-identical); the shipped .env opts in (BOB_IDENTITY_ENABLED).

Changed

• Removed the preview web UI — the desktop app is the GUI. The Preact browser stopgap is retired; the gateway now serves the JSON + WebSocket API only.

Fixed

• Codex health_check is CLI-only — no longer strands a native-GPT face under a team health-walk when the LiteLLM proxy is down.

Full notes in CHANGELOG.md. Reviewed by three independent models (Claude / GPT / Gemini) — no release-blocker. Manual gates for a full local run: KMM desktop visual check, native-Chinese translation review, and the LiteLLM-via-Codex worker path.

BoB v0.96.0 - first public release

Build - Orchestrate - Bind - a self-hosted, single-operator multi-agent orchestration platform. Your machine, your keys; its job is everything past aggregating models: routing work to the right model, fanning it out across many, deliberating in a council, and - the part most tools skip - verifying the result before it reaches you.

Headless-first (CLI / MCP / agent is the usable front door). The web and Kotlin Multiplatform desktop GUIs ship as a preview. Single-operator, loopback by default - the gateway is exposable for trusted remote access behind a TLS-terminating reverse proxy (read SECURITY.md first). Windows-first install.

Get started

git clone https://github.com/tyoung515-svg/bob.git
cd bob
./install-bob.ps1

See README.md to get going, ARCHITECTURE.md for the operating model, and SECURITY.md before exposing anything. Full notes in CHANGELOG.md.

Apache-2.0. A Canopy Seed project.