Bellamente v0.1.0

What's Changed

• Release v0.0.1: dev -> prod by @Jeff-Kazzee in #25
• Release v0.0.1 refresh: zero-config auth + roadmap by @Jeff-Kazzee in #27
• prod: Pages workflow by @Jeff-Kazzee in #30
• prod: Vercel hosting + roadmap fix by @Jeff-Kazzee in #32
• prod: docs site + downloads + changelog by @Jeff-Kazzee in #34
• prod: ship the docs pages (gitignore fix) by @Jeff-Kazzee in #82
• Release: retrieval quality wave — recency, MMR, temporal validity, eval harness, export/import by @Jeff-Kazzee in #95
• chore: promote v0.0.2 to prod by @Jeff-Kazzee in #111
• docs: update release truth instructions by @Jeff-Kazzee in #112
• docs: promote release truth follow-up by @Jeff-Kazzee in #113
• release: promote 0.1.0 to prod by @Jeff-Kazzee in #146
• fix(release): Node 22 in release workflow (prod hotfix for 0.1.0 publish) by @Jeff-Kazzee in #149

Full Changelog: v0.0.2...v0.1.0

Bellamente v0.0.2

Bellamente v0.0.2

Bellamente now has package-manager launchers in addition to direct binaries:

npm install -g bellamente
pipx install bellamente
uvx bellamente doctor

What Changed

• Added npm and PyPI launcher packages for bella.
• Added a package/release artifact gate to CI: npm pack dry-run, PyPI build + twine check, cross-platform binary build, and checksum validation.
• Moved public repo/site surfaces to The Little AI Company org and GitHub Pages.
• Kept the full-functionality smoke gate: real embedder, real database boot, buffered and streamed memory loops, export/import, temporal recall, and hard-delete scope.

Integrity Notes

• Launchers download the matching GitHub release binary and verify it against SHA256SUMS.txt before execution.
• Cached binaries are re-used only when their local .sha256 sidecar matches the binary.
• Windows binaries are unsigned in this early release; SmartScreen may warn even when checksums match.

2026-07-05: macOS binaries withdrawn

The bella-darwin-arm64 and bella-darwin-x64 assets were removed from this release: they never had a real test pass, and Bellamente only claims tested Windows x64 and Linux x64 binaries. Their lines remain in SHA256SUMS.txt so earlier downloads can still be verified. macOS binaries will return when a real test pass exists.

Also for the record: v0.0.2 first published the retrieval upgrades that landed on dev after v0.0.1 — memory full-text search, recency-weighted ranking, MMR diversity, temporal validity windows (asOf recall), export/import, and the route-level retrieval eval harness.

Bellamente v0.0.1 — Memoria Viva for your AI agents

Bellamente v0.0.1 — Memoria Viva for your AI agents

First release. Bellamente is a local-first memory tool for AI agents: it stores durable facts and
documents, recalls them semantically, and sits in front of your local LLM as a drop-in
/v1/chat/completions proxy that injects relevant memory — with a durable trace of exactly what it
did. One binary. No Docker, no account, no cloud.

Highlights

• Memory you can inspect and trust. Every recall is logged: what was searched, what matched, at
  what score, in how many milliseconds — readable via /inspect and the built-in dashboard.
  Corrections are versioned (nothing silently overwritten), forgetting is reversible and audited.
• Drop-in proxy for any OpenAI-compatible client. Change one base URL. Works with Ollama,
  LM Studio, llama.cpp, vLLM. Memory grounding runs on buffered and streamed requests — the
  proxy intercepts the model's searchMemory call (even mid-stream, even after preamble text),
  runs local recall, re-invokes your model, and streams only the final answer.
• Auto-capture with LLM distillation. Durable first-person facts from your chats are remembered
  through the standard dedup path — extracted by a small LLM pass through the SAME local upstream
  (never a cloud call), with heuristic fallback, a sensitive-content exclusion list
  (credentials/financial/medical are never stored), and a capture trace for every event.
  BELLA_PROXY_CAPTURE=0 disables; BELLA_CAPTURE_DISTILL=0 keeps capture heuristics-only.
• Embedded everything. Postgres + pgvector compiled to WASM run in-process (no server);
  embeddings are local and device-scaled (multilingual-e5-small on capable machines, a static
  Model2Vec model on low-RAM machines — pinned per data dir so hardware changes never corrupt
  recall). Append-only migrations mean upgrades never strand your data.
• Local-first by default. Binds 127.0.0.1 unless you opt out. No telemetry. Your machine is
  the cloud.

Also in this release

• Document ingestion (POST /documents): structure-aware markdown chunking, token-budget guarded;
  hybrid document search (vector + full-text, RRF-fused).
• Full memory lifecycle API: dedup/supersede on write, version-chain reads, PATCH-as-new-version,
  reversible forget, hard delete.
• Upstream timeouts + graceful degradation: a hung model server or failed recall never hangs or
  500s your chat turn.
• Dashboard: traces, search playground, memory browser with edit/forget/delete/history — styled to
  match the website ("La Macchina").

Install

Download the binary for your platform below, chmod +x (macOS/Linux), and run:

bella doctor   # verify: DB, model, ports, disk
bella          # serve on 127.0.0.1:8080

Point your client at http://127.0.0.1:8080/v1. Zero config: no .env, no API key on
loopback (the default). Exposing beyond loopback (BELLA_HOST) auto-generates a key in the data
dir and requires it; BELLA_API_KEY overrides. All other BELLA_* env vars are optional tuning.

Checksums: SHA256SUMS.txt.