v0.4.0

What's Changed

• AI spend limits in dollars, per person and per company by @blockgroot in #67

Full Changelog: v0.3.0...v0.4.0

v0.3.0

What's Changed

• Feat/deployment setup by @blockgroot in #60
• chunks: one partition per organization, row-level security on top by @blockgroot in #61
• Rate limiting: token buckets in Redis, per person and per company by @blockgroot in #65
• Each company's own Google and Slack app by @blockgroot in #64
• Exact answers from large, multi-tab sheets: every tab, worked figures, routing and chart fixes by @blockgroot in #66

Full Changelog: v0.2.1...v0.3.0

v0.2.1

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• Feature/signup approval queue v2 by @blockgroot in #1
• feat: enhance GitHub integration and UI updates by @blockgroot in #3
• Feature/GitHub integration by @blockgroot in #2
• fix(ingest): speed up onboarding sync and stop sticky Update by @blockgroot in #5
• feat(github): workspace-scoped GitHub connections by @blockgroot in #4
• feat(api): implement member removal and connection disconnection feat… by @blockgroot in #6
• Feature/jina remote embed rerank by @blockgroot in #7
• feat(email): add SendGrid support for email delivery by @blockgroot in #8
• feat(embeddings): implement batching for remote embedding provider an… by @blockgroot in #9
• fix(notion): bound block-tree fetch size to stop OOM on deep/wide pages by @blockgroot in #10
• feat(ingest): defense-in-depth memory guard + bounded SymSpell cache by @blockgroot in #11
• fix(ingestion): defer transformers import to first use, not module load by @blockgroot in #12
• fix(onboarding): a later sync must not revoke ready_to_ask by @blockgroot in #13
• fix(workspaces): don't show the owner a member-only view while loading by @blockgroot in #14
• Fix/end to end latency audit by @blockgroot in #15
• fix(ingestion): retry first sync once when Notion's index lags a fresh OAuth grant by @blockgroot in #16
• feat(slack): Phase 1 — Slack bot-token OAuth connect by @blockgroot in #17
• feat(slack): channel member management for workspace invites by @blockgroot in #18
• frontend: generalize policy-specific copy, add data privacy messaging by @blockgroot in #19
• Frontend/next updates by @blockgroot in #20
• Feature/linear integration by @blockgroot in #21
• feat: enhance frontend integration with Linear and update document by @blockgroot in #22
• feat: enhance setup status and chat interface with legacy document by @blockgroot in #23
• Prod optimisation by @blockgroot in #24
• Feature/drive pdf docx ingestion by @blockgroot in #25
• Feature/prompt driven scheduler by @blockgroot in #27
• Feat/multi model selection by @blockgroot in #28
• Fix/reasoning token starvation by @blockgroot in #29
• Feat/bring your own model by @blockgroot in #30
• Feature/auto sync freshness by @blockgroot in #31
• Fix/autosync skips GitHub by @blockgroot in #32
• Feature/space ux and copy by @blockgroot in #33
• fix(schedulers,workspaces): say who receives a report, and drop a tab… by @blockgroot in #34
• docs(plans): end-to-end plan for visual representation across every c… by @blockgroot in #35
• Infer recency intent from questions; boost/filter retrieval by date by @blockgroot in #36
• Add a LettuceDetect backend for the answer audit by @blockgroot in #41
• Plan: live connector access for the Second Brain (live-tools gateway) by @blockgroot in #43
• Second Brain: knowledge graph + live tool access + personal memory by @blockgroot in #39
• Close the missing-features gaps: per-document access everywhere, instant updates for all four sources by @blockgroot in #44
• Docs/product status by @blockgroot in #46
• Inline citations: [n] chips that open the source document by @blockgroot in #48
• feat(charts): open-ended charts, Chart mode, document tables, figures from text by @blockgroot in #45
• fix(charts): real names for focus, honest badges, no lone-bar trends by @blockgroot in #49
• Feat/benchmark suite by @blockgroot in #47
• fix(charts): count a merge the list omitted, and check the named repo by @blockgroot in #50
• Implement chart rendering for Slack and enhance document handling by @blockgroot in #51
• Fix/charts docs prose slack by @blockgroot in #52
• Deep analysis: answer high-level (company-wide) questions by @blockgroot in #53
• Plan: Act mode for write actions, with guardrails by @blockgroot in #54
• Small fixes: question-check cap, LangGraph warning, Slack diagnosis, follow-up reuse, Claude thinking per model by @blockgroot in #55
• Agent mode: chains continue after approval; a refused card can be corrected by @blockgroot in #58
• Mask personal identifiers before text leaves for an outside service by @blockgroot in #57
• Agent mode: fetched before drafted; a company's own model is tested f… by @blockgroot in #59

New Contributors

• @blockgroot made their first contribution in #1

Full Changelog: https://github.com/blockgroot/RAG-Engine/commits/v0.2.0