From 54570066e3c21edabe3a984d012b040df2f4f766 Mon Sep 17 00:00:00 2001 From: Ilker Manap Date: Thu, 8 Oct 2026 16:22:07 +0200 Subject: [PATCH] v0.14.0 Co-Authored-By: Claude Opus 5.5 (1M context) --- CHANGELOG.md | 90 ++++++++++++++++++++++++++++++++-------------------- 1 file changed, 55 insertions(+), 35 deletions(-) diff --git a/CHANGELOG.md b/CHANGELOG.md index 4c1354d..a3098c5 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -18,47 +18,67 @@ is only released once its **exit criterion** is met, not when its code lands. | v0.11.0 | — | High availability from setup, the setup and console screens, usage sums and per-tenant alerts | released | | v0.12.0 | — | The cluster's settings in the store, in force when changed; the firewall switch; wipe | released | | v0.13.0 | — | The whole system stopped and started from one command; Qwen3Guard judges, on a GPU and CPUs together | released | +| v0.14.0 | — | Knowledge bases: documents from people's own machines, kept apart per tenant, answered from | released | | v1.0.0 | 5 | Tenancy, prefix affinity, benchmarking | planned | Until v1.0.0 the minor version is the breaking one: configuration formats and Go APIs change between minors without a deprecation period. -## Unreleased +## v0.14.0 — 2026-10-08 -- **Knowledge bases, the store's part** (BACKLOG F1, branch `rag`): - collections of documents owned by an organization, team, project or - person, grants that add up down the tenancy tree and never cross an - organization, chunks in PostgreSQL with pgvector — partitioned by - organization, an HNSW index per collection, row-level security on the - table as a second line behind every query's own filter — and a search - that fuses nearest-neighbour and full-text ranks. Off on SQLite, and on - PostgreSQL without pgvector, saying why. `server init` and the - high-availability database setup install pgvector and enable it; a - cluster set up before needs `postgresql--pgvector` on every - database machine and `CREATE EXTENSION vector` in llmgw's database. A - project, team or organization that owns collections is not deleted; a - person's collections go with their account. -- **Knowledge bases filled from people's own machines** (BACKLOG F2): - `llmgw-kb`, its own download for Windows, macOS and Linux, reads documents - where they are, cuts them into chunks under their headings, has the - gateway embed them with the person's key, and sends text and vectors — - never the files — for what changed only, remembered in a SQLite ledger on - the machine. It reads PDF, Word, Excel, PowerPoint, OpenDocument, RTF, - EPUB, HTML, MHTML, mail with attachments, notebooks, subtitles, ZIP - archives and text and source in any common encoding; legacy Office files - with LibreOffice and scanned pages with Tesseract when installed. - `--dry-run` says what would leave the machine; `--delete` removes only - what that machine sent. The gateway takes them at `/v1/kb/…` with the - key's own access; administrators make, grant and search collections with - `llmgw ctl kb`, `/api/v1/kb` and the console's Knowledge screen - ([docs/knowledge.md](docs/knowledge.md)). -- **Answers from knowledge bases** (BACKLOG F3): a chat request naming - collections in `X-Llmgw-Knowledge` has its last question embedded on the - caller's account, the best passages found — by meaning, plus a little for - the question's words — and added to the system message, numbered, with - `x-llmgw-sources` saying which; before the guardrails judge the request, - never from the answer cache, and as 404 for a collection the key may not - read. Through the Responses and Messages APIs too. +Knowledge bases: a company's documents, put in by people from their own +machines, kept apart per tenant, and answered from. + +**Upgrading.** +- **pgvector on a cluster set up before this release.** Knowledge bases + need pgvector, which `server init` and the high-availability database + setup now install, but a rerun on a running cluster does not: + 1. Install `postgresql--pgvector` on every database machine. + 2. As postgres, run `CREATE EXTENSION vector` in llmgw's database. + 3. Restart the control planes. + Until then everything else works, and `llmgw ctl kb list` says why + knowledge bases are off. On a SQLite store they stay off. +- **An embedding model.** A collection needs a catalog model with + `capabilities.embeddings`, for example bge-m3, whose weights are in the + mirror. +- **The client.** People download `llmgw-kb` for their system from this + release. + +### Added + +- **Collections and who may read them.** A collection is owned by an + organization, team, project or person. Grants add up down the tenancy + tree and never cross an organization. A collection a key may not read is + answered as one that does not exist. Chunks are kept in PostgreSQL with + pgvector, partitioned by organization, with an HNSW index per collection, + and row-level security on the table as a second line behind every + query's own filter. A project, team or organization that owns collections + is not deleted; a person's collections go with their account. +- **`llmgw-kb`**, its own download for Windows, macOS and Linux, run by + people with their own key. It reads documents where they are and sends + text and vectors, never the files, for what changed only; a SQLite ledger + on the machine remembers what it did. + - It reads PDF, Word, Excel, PowerPoint, OpenDocument, RTF, EPUB, HTML, + MHTML, mail with attachments, notebooks, subtitles, ZIP archives, and + text and source code in any common encoding. + - Legacy Office files are read with LibreOffice, and scanned pages with + Tesseract, when those are installed. + - `--dry-run` says what would leave the machine; `--delete` removes only + what that machine sent. +- **Answers from documents.** A chat request names collections in + `X-Llmgw-Knowledge`, or its part of the tree has a default (`llmgw ctl kb + default`). The best passages, chosen by meaning plus a little for the + question's words, are added to the system message, numbered, and + `x-llmgw-sources` says which were used. This happens before the + guardrails judge the request, and the answer is never served from the + answer cache. On vLLM replicas the organization's `cache_salt` keeps + tenants' prefix caches apart. It works through the Responses and Messages + APIs too. +- **Administration**: `llmgw ctl kb` (create, list, grant, revoke, default, + docs, rm, search, delete), `/api/v1/kb`, the console's Knowledge screen, + and a Documents picker on the portal's chat page. +- Metrics `llmgw_kb_retrievals_total` and `llmgw_kb_retrieval_seconds`; + [docs/knowledge.md](docs/knowledge.md). ## v0.13.0 — 2026-10-08