Skip to content

CHANGELOG.md

Latest commit

 

History

History
363 lines (296 loc) · 31.4 KB

CHANGELOG.md

File metadata and controls

363 lines (296 loc) · 31.4 KB

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

[1.8.5] - 2026-04-07

Changed

  • Compressed the public front door again so the first screen keeps the product center on messy long context -> structured insight -> inspectable outcomes, while leaving MCP, starter bundles, distribution, and promotion surfaces in the second ring.
  • Clarified the docs/front-door routing so README -> quickstart -> proof is the shortest honest evaluation order, and promotion/status pages stop competing with the first product read.
  • Split CI governance more cleanly into pre-commit / pre-push / hosted / nightly / manual, moved lingering weekly maintenance language to nightly, and made the default local pre-push rehearsal path fast by default while keeping full as the stricter opt-in path.
  • Refreshed the active task-board and required-check snapshots with 2026-04-07 live truth so older same-day release/homepage claims no longer masquerade as current state.
  • Raised the managed Python security floor by pinning pillow>=12.1.1,<13 through tool.uv.override-dependencies, narrowed uv lock resolution to the current Linux/macOS support surface, and refreshed both uv.lock files so the in-repo dependency graph no longer requires the vulnerable pillow < 12.1.1 line for active supported environments.
  • Added repo-owned security/governance workflows for GitHub Dependency Review, Trivy, TruffleHog, and a scoped Zizmor lane, while refreshing the pinned GitHub Actions dependency SHAs for checkout, setup-node, setup-python, github-script, artifact transfer, Docker Buildx/build-push, and the Claude Code action.
  • Hardened the composite Python bootstrap action against template-injection by parsing sync-args through a shell-safe argv bridge instead of interpolating raw workflow input directly into the run_uv_managed.sh command line.
  • Converted the stale public-distribution task-board snapshot into an explicitly non-authoritative historical artifact by recording the 2026-04-06 closeout addendum: main/origin/main = 9d2a4a6, open PRs = 20, draft releases = [], open_graph_image_url = null, open code/secret scanning alerts = 0, and the two active pillow Dependabot alerts targeted in this wave.
  • Added an explicit Dependabot ignore ledger for the currently unlandable apps/web and evals/promptfoo update wave so unsupported or unverified package bumps stop reopening as if they were silently accepted.
  • Aligned the app root first door with the documented source-first product path by redirecting / to /sources instead of the notebook lane, so first-run entry now matches the repo's long-context collection story.
  • Hardened local runtime host-process safety with fail-closed PID-record handling in the start_*_local.sh entrypoints, a new host-process safety CI/runtime gate, and contract coverage that forbids broad kill/desktop-automation primitives in tracked repo automation.
  • Required an explicit PROVENOTE_ALLOW_DETACHED_CHROME_LAUNCH=1 operator override before detached repo-owned Chrome launches can start from the real-profile helper, keeping manual browser control inside an ownership-first boundary.
  • Tightened the tracked OpenClaw example-bundle skills so they preserve the same read-first validation loop and non-public-boundary wording as the repo-local provenote-mcp-outcome-workflows skill, instead of drifting into over-broad compatibility claims.
  • Removed the broken cache-to: type=gha export path from both Development Build and Build and Release; those workflows still read from shared GHA cache scopes, but no longer let optional cache export turn a successfully built GHCR image into a red build.
  • Aligned the frontend test toolchain after the @vitest/ui bump by moving vitest, @vitest/coverage-istanbul, and @vitest/coverage-v8 onto the same 4.1.2 family so fresh installs stop failing peer resolution.
  • Refreshed the frontend dependency surface with @tailwindcss/postcss 4.2.2 and lucide-react 1.7.0, while replacing the removed YouTube icon export in SourceContentTab with a stable play glyph so the existing content flow keeps rendering.
  • Fixed the Cursor-style host bundle contract test so it now checks the bundle command file as a tracked, non-ignored Git path instead of relying on git status --untracked-files=all for a clean tracked file.
  • Refreshed the closeout truth artifacts so the active plan/task-board layer no longer claims older same-day SHA snapshots or a zero-open-PR state that no longer matches live GitHub truth.
  • Clarified the ecosystem truth boundary across the public docs surface so host guides, the first-party CLI, public-skills non-claims, OpenClaw defer status, plugin/marketplace non-claims, and owner-only external actions stay in separate buckets.
  • Tightened front-door audience routing across the README, proof page, and quickstart so messy-context, research, coding-agent, operator, and readiness readers can reach the right proof path without mistaking host pages for the product center.
  • Refreshed the public status and brand/domain boundary pages with the current external-window framing: repo metadata can be ready before release publication, domain moves, trademark decisions, listings, or partnership actions happen.
  • Hardened the outcome client typing surface and the MCP research_thread.create typed envelope so runtime type gates pass without changing the current outcome-lane behavior.
  • Fixed the runtime lint gate so template-only directories no longer get misclassified as Python source roots during pre_commit_lint.sh --mode runtime.
  • Added a strict-health option to provenote status --json so operator and agent workflows can fail fast when the local /health probe is down instead of only printing a degraded payload.
  • Brought the Claude Code, Codex, and Cursor integration pages up to the same repo-backed proof-loop standard as the OpenCode page, so host compatibility can be verified from repo-owned entrypoints and tests instead of resting on thin setup copy alone.
  • Aligned both Docker runtime build paths with the app-local frontend artifact contract, so ops/docker/Dockerfile and ops/docker/Dockerfile.single now copy .runtime-cache/build/next/standalone and .runtime-cache/build/next/static instead of stale .next assumptions.
  • Added tracked Claude Code, Codex, Cursor, and OpenCode starter bundles under examples/hosts/, added a dedicated OpenClaw local proof/prep page, and tightened the OpenClaw-compatible bundle family so local install artifacts, skills, and MCP config stay repo-owned without drifting into plugin-store or marketplace claims.
  • Hardened registry-auth and release-proof truth around the host bundle and image lanes: release proof now checks explicit GHCR auth linkage, and Development Build degrades honestly when GHCR push access is unavailable instead of turning a successful build into a misleading red publish failure.
  • Surfaced the tracked host starter bundles from the public front door so developers can reach repo-owned copy/install artifacts directly from the README instead of discovering them only through deeper docs.
  • Fixed the Gemini startup probe so SDK-thrown model-404 errors now normalize into the existing fallback semantics instead of crashing startup before the stable fast-path probe can continue.
  • Added a host-example landing guard for the Cursor bundle so .cursor/commands/provenote-mcp-outcome-workflows.md stays visible to Git and contract tests instead of being silently swallowed by the root .gitignore .cursor/ rule.

[1.8.4] - 2026-04-01

Changed

  • Finalized the local-to-remote closeout wave: the queue-burst rollout now converges as a source-grounded knowledge-work control tower with synchronized README, docs index, project-status, proof, and compatibility surfaces.
  • Promoted the previously local-only closeout truth to committed and pushed main, including the authoritative version-grade handoff package and refreshed task board / closeout artifacts.
  • Completed the high-value journey i18n convergence pass: interpolated notebook/source/research journey strings now flow through the shared locale contract, test translation mocks resolve placeholders correctly, and the main journey frontend suites are green again.
  • Kept API, MCP, generated client, and builder-facing wording aligned under the same product truth: first-party MCP remains the integration surface, not the brand center, and current public wording stays compatibility-first without official-partnership overclaim.
  • Refreshed cleanup evidence with a new audit/apply/recheck pass, including repo-local runtime/cache hygiene, machine-cache cleanup, and final __pycache__ removal while preserving protected local state.

[1.8.3] - 2026-03-28

Changed

  • The release workflow now refuses to publish a release event from an out-of-date tag SHA and only uploads release-proof artifacts when the corresponding image build actually succeeds, reducing noisy false downstream failures in Build and Release.
  • Release publication semantics are now split cleanly: the real image publication path stays on workflow_dispatch from the trusted main head, while the release event only verifies that a same-SHA successful manual build already exists before allowing the public release page wave to pass.
  • Current fork drift governance now uses a refreshed live_git_truth sample with direct repo-side validation, and no-merge-base topology is now treated as a first-class selective-port reality instead of an ordinary merge/rebase drift story.
  • Public-facing repository wording now describes the current path as a repo-documented local proof loop rather than a one-minute hosted trial, and release visibility is explicitly separated from release health in repo-tracked public surface snapshots.
  • Active runtime source and prompt defaults now keep English-only canonical behavior, with a dedicated runtime English boundary gate and contract test wired into the deterministic CI path.
  • The tracked public asset pool is now explicitly guarded in CI, and the broken quick-result storyboard GIF has been removed from the repository asset surface.

[1.8.2] - 2026-03-28

Changed

  • This release tag captured the first governance closeout wave but did not become the final clean public release proof because its release-event build was published before the new same-SHA timing model had been fully exercised.

[1.8.1] - 2026-03-25

Changed

  • Public-facing repository narrative was rewritten around a product-first Provenote identity: the README now leads with result path, proof map, why-star framing, and layered docs entrypoints instead of opening with governance-heavy boundary language.
  • Public docs were reorganized for first-time visitors: docs/index.md now routes by user goal, and new docs/quickstart.md, docs/proof.md, and docs/faq.md provide evaluation, onboarding, and public-proof entry surfaces.
  • Public visual assets were added for GitHub-native growth and sharing, including a hero visual, proof-stack visual, architecture visual, and a social preview image under docs/assets/.
  • GitHub growth surfaces were tightened around Provenote's current public identity, including repository description/topics/discussions strategy, release-note category configuration, and discussion-first support routing in the issue chooser.
  • Auto-assign model policy now tolerates provider naming aliases (for example gemini-3-flash-preview vs gemini-3.0-flash) in services/services/api/routers/models.py, and related API regression tests were aligned with current fail-closed idempotency and source-processing error semantics to keep local preflight deterministic.
  • UIUX gate hardening for shared self-hosted runners: Playwright browser install now retries with backoff in .github/workflows/uiux-gemini-gate.yml to tolerate transient apt lock contention (/var/lib/apt/lists/lock) under concurrent CI load.
  • UIUX visual baseline stability calibration: increased maxDiffPixels for Chromium snapshots in search/settings/sources E2E visual-baseline tests while keeping maxDiffPixelRatio=0.03, reducing cross-runner rendering false negatives without relaxing ratio-based strictness.
  • Playwright CI artifact policy tightened for self-hosted disk stability: screenshot capture is now only-on-failure (instead of on) while keeping trace/video retry evidence, reducing ENOSPC risk without weakening failure diagnostics.
  • Backend coverage merge reliability: normalized cross-runner coverage path handling via tool.coverage.run.relative_files + tool.coverage.paths and made test/merge steps explicitly use pyproject.toml coverage config, fixing No source for code failures in Tests workflow coverage aggregation.
  • Mutation profile config cleanup: removed three intermediate mutants/tests/test_mutation_survivor_killers.py injections from pyproject.toml; runtime survivor-killer selection now stays script-driven via tests/test_mutation_survivor_killers_runtime.py.
  • Mutation profile execution now uses repo-relative survivor-killer test selection (mutants/tests/test_mutation_survivor_killers.py) instead of runtime-generated absolute temp paths, fixing mutmut collect failures on workspaces whose absolute path includes [].
  • Frontend UI accessibility hardening: clickable notebook/source cards are now keyboard-activatable (Enter/Space) with explicit interactive semantics, source card status/progress colors now follow theme tokens, and Playwright a11y gate now scans /notebooks, /sources, /settings in both light and dark themes.
  • CI queue-pressure optimization: Pre-commit now runs on push only for main (while still running on pull_request), Claude Code interaction workflow runs on ubuntu-latest, heavy execution lanes are pinned to e2-core, and build workflows (Development Build, Build and Release) enforce workflow-level concurrency to prevent redundant queued runs.
  • Additional queue hardening: maintenance/control workflows (JSCPD Duplication Check, Pre-commit Outdated Check, Upstream Drift Check, Claude Code Review, UIUX Auto Remediation) are now hosted-only with explicit workflow-level concurrency to preserve e2-core slots for heavy trusted lanes.
  • Build acceleration hardening: Development Build and Build and Release switched Docker Buildx cache from local filesystem cache to shared type=gha scopes, improving cache reuse across self-hosted runners while preserving existing release/quality gates.
  • Billing-fallback routing: when GitHub-hosted jobs are unavailable due billing limits, PR gate workflows (Pre-commit, JSCPD Duplication Check, Upstream Drift Check, Claude Code Review, Auditable Quality Gate, and Tests) now execute on self-hosted,e2-core to prevent immediate startup failures.
  • CI reliability hardening on self-hosted: Pre-commit now auto-cleans corrupted cache and retries once, heavy change-detection workflows explicitly grant pull-requests: read for dorny/paths-filter, and Claude Code Review is gated by vars.CLAUDE_CODE_REVIEW_ENABLED to avoid failing when the review token is not configured.
  • Drift governance scope update: Upstream Drift Check is now manual-only (workflow_dispatch) so selective-port review stays available without turning every main merge into a predictable red maintenance lane.
  • Release-proof GHCR auth fix: tooling/scripts/ci/export_oci_evidence.py now uses workflow-provided GHCR credentials when available, so same-SHA witness export no longer depends on anonymous GHCR pulls when the repository-scoped package remains private.
  • Development Build shared-runner hardening: privileged disk cleanup now degrades to best-effort when passwordless sudo is unavailable, and build jobs use isolated Docker config state so macOS keychain prompts do not break Buildx bootstrap before dev image publication even starts.
  • Single-container release fix: ops/docker/Dockerfile.single now uses a Docker-compatible frontend stage name (apps-web-builder), unblocking the single-image build path used by both Development Build and Build and Release.
  • UIUX gate stability hardening: accessibility E2E now tolerates long-running route hydration in CI (domcontentloaded + bounded networkidle wait), and visual thresholds were calibrated for self-hosted Chromium variance without relaxing ratio guardrails.
  • Self-hosted resilience hardening: UIUX Gemini Gate now performs bounded runner disk hygiene before E2E (_diag/_temp retention cleanup + stale report cleanup) and prints disk telemetry to prevent "No space left on device" runner crashes.
  • Automatic hosted fallback switch: hosted maintenance workflows (Pre-commit Outdated Check, UIUX Auto Remediation, Claude Code) now support HOSTED_FALLBACK_TO_E2_CORE; a new self-hosted workflow (Hosted Runner Fallback Autoswitch) detects hosted quota/billing pickup failures, enables fallback, and reruns failed jobs on e2-core.
  • Spot burst control for heavy lanes: heavy CI jobs now support CI_HEAVY_USE_SPOT; enabling it routes mutation/E2E/live-integration heavy tasks to self-hosted,e2-core,spot, keeping default behavior stable while unlocking burst capacity when spot runners are available.
  • Funnel routing hardening: lightweight CI control lanes now default to ubuntu-latest with variable-driven autoswitch fallback (HOSTED_FALLBACK_TO_E2_CORE) while heavy lanes remain self-hosted-only (e2-core/spot) for deterministic throughput.
  • Fallback lane isolation hardening: hosted-fallback jobs now target self-hosted,e2-core-dedicated (core-only) instead of broad e2-core, preventing lightweight fallback traffic from stealing spot burst slots reserved for heavy CI lanes.
  • Strict key/env fail-fast gate: added tooling/tooling/scripts/ci/check_required_ci_env.sh and wired it into critical workflows (Pre-commit, JSCPD, Tests, Auditable Quality Gate, UIUX Gemini Gate) so missing/placeholder GEMINI_API_KEY or OPEN_NOTEBOOK_ENCRYPTION_KEY fails CI immediately with no bypass path.
  • Local resource policy update: pre-push hook set is slimmed to fast governance checks; heavy static/runtime audits are CI-enforced to keep developer machines responsive without relaxing merge gates.
  • Pre-commit load rebalancing: expensive governance scans (test-smells, python test-smells, live static audit, env/secret governance, navigation docs pair, Gemini UIUX semantic audit, workflow-policy contract tests) were moved from pre-commit to pre-push, keeping commit-time feedback fast while preserving strict push/CI enforcement.
  • Workflow policy hardening (tooling/tooling/scripts/ci/check_workflow_policy.py): now additionally enforces strict env gate wiring (check_required_ci_env.sh + both required secrets) and codified runner routing contracts for hosted fallback lanes vs heavy self-hosted/spot lanes.
  • Live workflow secret parity: live-integration.yml now uses the same strict required-secret validator (check_required_ci_env.sh) as other critical workflows instead of a simple non-empty key check.
  • Heavy runner routing parity: Tests backend shards/frontend coverage and Auditable Quality Gate eval/property jobs now support CI_HEAVY_USE_SPOT (self-hosted,e2-core,spot) with stable fallback to self-hosted,e2-core.
  • Heavy lane routing consistency: test.yml property-tests now also supports CI_HEAVY_USE_SPOT, and workflow policy contracts were updated accordingly to keep heavy lanes uniformly spot-capable.
  • Local hook cost trimming: gemini-uiux-audit-pre-commit is moved to manual stage (optional local execution), while strict UIUX quality gate remains mandatory in CI workflow.

[1.7.4] - 2026-02-18

Fixed

  • Embedding large documents (3MB+) fails with 413 Payload Too Large (#594)
  • generate_embeddings() now batches texts in groups of 50 with per-batch retry, preventing provider payload limits from being exceeded
  • 413 errors now classified with user-friendly message in error classifier
  • Misleading "Created 0 embedded chunks" log in process_source_command — embedding is fire-and-forget, so the count was always 0; now logs "embedding submitted" instead

[1.7.3] - 2026-02-17

Added

  • Retry button for failed podcast episodes in the UI (#211, #218)
  • Error details displayed on failed podcast episodes (#185, #355)
  • POST /podcasts/episodes/{id}/retry API endpoint for re-submitting failed episodes
  • error_message field in podcast episode API responses

Fixed

  • Podcast generation failures now correctly marked as "failed" instead of "completed" (#300, #335)
  • Disabled automatic retries for podcast generation to prevent duplicate episode records (#302)

Dependencies

  • Bump podcast-creator to >= 0.11.2
  • Bump esperanto to >= 2.19.4

[1.7.2] - 2026-02-16

Added

  • Error classification utility that maps LLM provider errors to user-friendly messages (#506)
  • Global exception handlers in FastAPI for all custom exception types with proper HTTP status codes
  • getApiErrorMessage() frontend helper that falls back to backend messages when no i18n mapping exists

Fixed

  • LLM errors (invalid API key, wrong model, rate limits) now show descriptive messages instead of "An unexpected error occurred" (#590)
  • SSE streaming error events in source chat and ask hooks were swallowed by inner JSON parse catch blocks
  • Transformation execution errors were caught and re-wrapped as generic 500s instead of using proper status codes
  • Fail fast when source content extraction returns empty instead of retrying (#589)
  • Chat input and message overflow with long unbroken strings (#588)
  • Word-wrap overflow in source cards, note editor, inline edit, note titles, and dialog content (#588)
  • Translation proxy shadowing name keys (#588)
  • OpenAI-compatible provider name handling via Esperanto update (#583)

Changed

  • ValueError replaced with ConfigurationError in model provisioning for proper error classification
  • ConfigurationError added to command retry stop_on lists to avoid retrying permanent config failures

Dependencies

  • Bump esperanto to 2.19.3 (#583)
  • Bump podcast-creator to 0.9.1

[1.7.1] - 2026-02-14

Added

  • French (fr-FR) language support (#581)
  • CI test workflow and improved i18n validation (#580)
  • Expose embed command_id in note API responses (#545)

Fixed

  • ElevenLabs TTS credential passthrough via Esperanto update (#578)
  • Handle empty/whitespace source content without retry loop (#576)
  • Increase transformation max_tokens and update Esperanto dep (#568)
  • Turn the embedding field into optional (#557)

Docs

  • Fix docker container names in local setup guides (#577)

Dependencies

  • Bump langchain-core from 1.2.7 to 1.2.11 (#564)
  • Bump cryptography from 46.0.3 to 46.0.5 (#563)

[1.7.0] - 2026-02-10

Added

  • Credential-Based Provider Management (#477)

    • New Settings → API Keys page for managing AI provider credentials via the UI
    • Support for 14 providers: OpenAI, Anthropic, Google, Groq, Mistral, DeepSeek, xAI, OpenRouter, Voyage AI, ElevenLabs, Ollama, Azure OpenAI, OpenAI-Compatible, and Vertex AI
    • Secure storage of API keys in SurrealDB with field-level encryption (Fernet AES-128-CBC + HMAC-SHA256)
    • One-click connection testing, model discovery, and model registration per credential
    • Migration tool to import existing environment variable keys into the credential system
    • Azure OpenAI support with service-specific endpoints (LLM, Embedding, STT, TTS)
    • OpenAI-Compatible support with per-service URL configurations
    • Vertex AI support with project, location, and credentials path
    • Environment variable API keys deprecated in favor of Settings UI

  • Security Enhancements

    • Docker secrets support via _FILE suffix pattern (e.g., OPEN_NOTEBOOK_PASSWORD_FILE)
    • Default encryption key derived from "0p3n-N0t3b0ok" for easy setup (change in production!)
    • Default password "open-notebook-change-me" for out-of-box experience (change in production!)
    • URL validation for SSRF protection - blocks private IPs and localhost (except for Ollama which runs locally)
    • Security warnings logged when using default credentials

  • HTML clipboard detection for text sources (#426)

    • When pasting content, automatically detects HTML format (e.g., from Word, web pages)
    • Shows info message when HTML is detected, informing user it will be converted to Markdown
    • Preserves formatting that would be lost with plain text paste
    • Bump content-core to 0.11.0 for HTML to Markdown conversion support

  • Improved Getting Started Experience

    • Simplified docker-compose.yml in repository root (single official file)
    • Added examples/ folder with ready-made configurations:
      • docker-compose-ollama.yml - Local AI with Ollama
      • docker-compose-speaches.yml - Local TTS/STT with Speaches
      • docker-compose-full-local.yml - 100% local setup (Ollama + Speaches)
    • Inline quick start in README (no need to navigate to docs)
    • Cross-references between docker-compose examples and documentation
    • .env.example template with all configuration options

Fixed

  • Azure form race condition: all configuration now saved in single atomic request
  • Migration API "error error" display: added proper MigrationResult model with message field
  • Connection tester for Ollama providers: improved error handling and URL validation
  • SqliteSaver async compatibility issues in chat system (#509, #525, #538)
  • Re-embedding failures with empty content (#513, #515)
  • Deletion cascade for notes and sources (#77)
  • YouTube content availability issues (#494)
  • Large document embedding errors (#489)

Security

  • API keys are encrypted at rest using Fernet symmetric encryption
  • Keys are never returned to the frontend, only configuration status
  • SSRF protection prevents internal network access via URL validation

Docs

  • Complete documentation update for credential-based system across 25 files
  • All quick-start, installation, and configuration guides now use Settings UI workflow
  • Environment variable API key instructions moved to deprecated/legacy sections
  • Fixed broken links in installation docs
  • Added comprehensive examples/ folder with documented docker-compose configurations
  • Updated local-tts.md and local-stt.md with links to ready-made examples

Internationalization

  • Added Russian (ru-RU) language support (#524)
  • Added Italian (it-IT) language support (#508)

[1.6.2] - 2026-01-24

Fixed

  • Connection error with llama.cpp and OpenAI-compatible providers (#465)
    • Bump Esperanto to 2.17.2 which fixes LangChain connection errors caused by garbage collection

[1.6.1] - 2026-01-22

Fixed

  • "Failed to send message" error with unhelpful logs when chat model is not configured (#358)
    • Added detailed error logging with model selection context and full traceback
    • Improved error messages to guide users to Settings → Models
    • Added warnings when default models are not configured

Docs

  • Ollama troubleshooting: Added "Model Name Configuration" section emphasizing exact model names from ollama list
  • Added troubleshooting entry for "Failed to send message" error with step-by-step solutions
  • Updated AI Chat Issues documentation with model configuration guidance

[1.6.0] - 2026-01-21

Added

  • Content-type aware text chunking with automatic HTML, Markdown, and plain text detection (#350, #142)
  • Unified embedding generation with mean pooling for large content that exceeds model context limits
  • Dedicated embedding commands: embed_note, embed_insight, embed_source
  • New utility modules: chunking.py and embedding.py in packages/core/utils/
  • Japanese (ja-JP) language support (#450)

Changed

  • Embedding is now fire-and-forget: domain models submit embedding commands asynchronously after save
  • rebuild_embeddings_command now delegates to individual embed_* commands instead of inline processing
  • Chunk size reduced to 1500 characters for better compatibility with Ollama embedding models
  • Bump Esperanto to 2.16 for increased Ollama context window support

Removed

  • Legacy embedding commands: embed_single_item_command, embed_chunk_command, vectorize_source_command
  • needs_embedding() and get_embedding_content() methods from domain models
  • split_text() function from text_utils (replaced by chunk_text() in chunking module)

Fixed

  • Embedding failures when content exceeds model context limits (#350, #142)
  • Empty note titles when saving from chat (clean thinking tags from prompt graph output)
  • Orphaned embedding/insight records when deleting sources (cascade delete)
  • Search results crash with null parent_id (defensive frontend check)
  • Database migration 10 cleans up existing orphaned records

[1.5.2] - 2026-01-15

Performance

  • Improved source listing speed by 20-30x (#436, closes #351)
    • Added database indexes on source field for source_insight and source_embedding tables
    • Use SurrealDB FETCH clause for command status instead of N async calls

[1.5.1] - 2026-01-15

Fixed

  • Podcast dialog infinite loop error caused by excessive translation Proxy accesses in loops
  • Podcast dialog UI freezing when typing episode name or additional instructions
  • Removed incorrect translation keys for user-defined episode profiles (user content should not be translated)

[1.5.0] - 2026-01-15

Added

  • Internationalization (i18n) support with Chinese (Simplified and Traditional) translations (#371, closes #344, #349, #360)
  • Frontend test infrastructure with Vitest (#371)
  • Language toggle component for switching UI language (#371)
  • Date localization using date-fns locales (#371)
  • Error message translation system (#371)

Fixed

  • Accessibility improvements: added missing id, name, and autoComplete attributes to form inputs (#371)
  • Added DialogDescription to dialogs for Radix UI accessibility compliance (#371)
  • Fixed "Collapsible is changing from uncontrolled to controlled" warning in SettingsForm (#371)
  • Fixed lint command for Next.js 16 compatibility (eslint instead of next lint)

Changed

  • Dockerfile optimizations: better layer caching, --no-install-recommends for smaller images (#371)
  • Dockerfile.single refactored into 3 separate build stages for better caching (#371)

[1.4.0] - 2026-01-14

Added

  • CTA button to empty state notebook list for better onboarding (#408)
  • Offline deployment support for Docker containers (#414)

Fixed

  • Large file uploads (>10MB) by upgrading to Next.js 16 (#423)
  • Orphaned uploaded files when sources are removed (#421)
  • Broken documentation links to ai-providers.md (#419)
  • ZIP support indication removed from UI (#418)
  • Duplicate Claude Code workflow runs on PRs (#417)
  • Claude Code review workflow now runs on PRs from forks (#416)

Changed

  • Upgraded Next.js from 15.4.10 to 16.1.1 (#423)
  • Upgraded React from 19.1.0 to 19.2.3 (#423)
  • Renamed middleware.ts to proxy.ts for Next.js 16 compatibility (#423)

Dependencies

  • next: 15.4.10 → 16.1.1
  • react: 19.1.0 → 19.2.3
  • react-dom: 19.1.0 → 19.2.3

[1.2.4] - 2025-12-14

Added

  • Infinite scroll for notebook sources - no more 50 source limit (#325)
  • Markdown table rendering in chat responses, search results, and insights (#325)

Fixed

  • Timeout errors with Ollama and local LLMs - increased to 10 minutes (#325)
  • "Unable to Connect to API Server" on Docker startup - frontend now waits for API health check (#325, #315)
  • SSL issues with langchain (#274)
  • Query key consistency for source mutations to properly refresh infinite scroll (#325)
  • Docker compose start-all flow (#323)

Changed

  • Timeout configuration now uses granular httpx.Timeout (short connect, long read) (#325)

Dependencies

  • Updated next.js to 15.4.10
  • Updated httpx to >=0.27.0 for SSL fix