Skip to content

AI Private Workspace v0.2.0

Choose a tag to compare

View all tags
@github-actions github-actions released this 24 Jun 12:46
v0.2.0
f3955de

Downloads

macOS

Windows

First minor release since 0.1.99. It turns AI Private Workspace from a local RAG
chat into a full read-only project-intelligence tool — still entirely local,
private, and grounded in your own files — and hardens the project's open-source
security posture. Headline themes (the detailed 0.1.100–0.1.178 history is
preserved below):

Added

  • Project Intelligence. A deterministic, role-neutral evidence graph with an interactive Map; role lenses (Developer / DevOps / Tester / Business analyst) and an adaptive role dashboard; a visual CI/CD flow; Cloud and References tabs; environment comparison and deployment flow; and a read-only Security lens. Findings everywhere read as human leads ("what was found, why it may matter, what to check"), not verdicts.
  • Read-only agents & analysis. The Investigator — a bounded ReAct loop over read-only tools (search_code, read_file, graph_query, list_files, git_history, ci_triggers) with a transparent trace and collected sources — and the deterministic Watcher ("what changed since I last looked?", with an optional one-tap LLM recap of the commits).
  • Project groups. Treat several repositories as one project: group Ask (repo-tagged sources), a portfolio Home, and group Intelligence that compares rather than merges (repo×environment matrix, shared-vs-unique tech, risks grouped by pattern).
  • Memory & profile. Local, fully-editable project memory + handbook, and a cross-project "About you" profile applied to every project — with review-first suggestion (nothing saved until you keep it).
  • Navigation. A Cmd/Ctrl-K command palette and a file inspector (owner, change-coupling, blast radius, the risks touching a file).
  • Answer feedback that teaches the project (👍/👎 → corrections fed back into prompts) and much richer git intelligence (activity over time, branch strategy, change coupling, merge/PR activity).
  • Embedders & models. BGE-M3 / Qwen3-Embedding installable on both engines; Qwen3 4B as the default answer model; conversation memory and an opt-in reranker in Ask.

Security & supply chain

  • OpenSSF Best Practices: passing, plus OpenSSF Scorecard, CodeQL SAST, REUSE-compliant licensing, and a CodeFactor grade. All GitHub Actions are pinned to commit SHAs; workflow tokens follow least privilege; the backend Docker image is pinned by digest. Releases are signed and ship SHA256SUMS, an SPDX SBOM, and an automated-test report. Contact routes through GitHub (no email in the repo).

Changed

  • The Skills library is now a single editor in Settings; the Models tab gains Tuning. Bundled llama.cpp updated to b9777.
All commits in this release
  • release: 0.2.0
  • chore(release): 0.1.178 — OpenSSF passing badge, Scorecard hardening, build fixes
  • fix(lint): resolve GitChangeBrief forward-ref annotations under TYPE_CHECKING (ruff F821)
  • fix(build): vite manualChunks as function form (type-compatible after Vite/Rollup bump)
  • docs(readme): add OpenSSF Best Practices passing badge (project 13357)
  • chore(deps): update httpx requirement in /backend
  • chore(deps): update pathspec requirement in /backend
  • chore(ci): bump actions/checkout from 6.0.2 to 7.0.0
  • build: bump backend Docker base image to python:3.14-slim (matches local), pinned by digest
  • chore(deps): bump the frontend-dev group in /frontend with 4 updates
  • chore(deps): bump time from 0.3.36 to 0.3.51 in /frontend/src-tauri
  • chore(ci): bump actions/upload-artifact from 4.6.2 to 7.0.1
  • chore(deps): update pydantic requirement in /backend
  • chore(ci): bump actions/setup-python from 6.2.0 to 6.3.0
  • chore(ci): bump github/codeql-action from 3.36.0 to 4.36.2
  • chore(deps): update qdrant-client requirement in /backend
  • chore(deps): update fastapi requirement in /backend
  • chore(deps): bump python from 3.11-slim to 3.14-slim in /backend
  • build(security): pin Docker base image by digest; add docker ecosystem to Dependabot
  • ci(security): least-privilege token perms in release.yml; link content in SECURITY.md
  • docs(readme): move CI to top row, license to the compliance row
  • docs(readme): use native REUSE + CodeFactor badges (shields proxy showed stale/unregistered)
  • docs(readme): group badges into two centered rows, unify badge style
  • chore: route contact through GitHub, drop email from copyright
  • chore(security): set up OpenSSF/REUSE/CodeQL badges + pin actions; bump llama.cpp b9777
  • feat(watch): one-tap LLM recap of commits since last check
  • feat(watch): git-grounded human change brief in 'What changed since last time'
  • fix(watch): correct plural in change digest (dependencys -> dependencies)
  • fix(routes): import get_settings in workspaces.py (NameError in understanding route)
  • feat(analysis): fit project-understanding prompt to the context window
  • fix(llamacpp): raise answer-model context to 8192 (was 4096)
  • fix(index): tag untagged index on embedder switch so reindex prompt is reliable
  • fix(index): show reindex prompt only when the search model changed
  • refactor(skills): single editor in Settings; remove Models skill duplicate
  • fix(models): embedding provider reports the engine's actual model on startup/open
  • chore(ui): About-you Style placeholder example -> 'Answer in Ukrainian'. 0.1.166
  • feat(brand): use light/dark pigeon icons for tray + in-app logo, theme-swapped
  • fix(models): re-point live embedding provider when switching search model
  • feat(desktop): system tray icon (mac menu bar + windows), theme-aware on macOS
  • feat(index): structure-aware chunking (Python AST + brace-aware + fenced markdown)
  • fix(setup): don't bounce to onboarding takeover when switching the search model
  • fix(ui): theme-aware in-app logo (dark logo in dark mode, light in light)
  • fix(models): refresh 'current setup' after switching llama.cpp answer/search model
  • fix(models): switch + delete the built-in search/embedding model
  • fix(models): re-attach to running llama.cpp download after panel remount
  • test: guided setup default is now qwen3:4b
  • feat(models): make new embedders installable on both engines + refresh default to Qwen3 4B
  • feat(memory): review-first profile-fact suggestions
  • chore(models): drop mxbai embedder suggestion + fix catalog count tests
  • ui(models): clearer at-a-glance embedder comparison + hover tip
  • feat(memory): user profile — cross-project memory about the person
  • feat(embeddings): add BGE-M3 + Qwen3-Embedding 0.6B to the catalog
  • ui(ask): remove the Assistant focus panel
  • ui(settings): open skills editor on the workspace's active role
  • fix(roles): role change also sets the active Ask skill
  • feat(roles): unify to 5 canonical roles + persist role as one workspace setting
  • feat(role): land on role's tab + mark 'for your role' risks
  • fix(role-brief): de-duplicate top risks by title in the dashboard band
  • fix(groups): drop deleted workspaces from group member counts
  • ui: drop duplicate risk badge, de-dupe Overview counts, fix header bleed
  • ui: calmer risk cards + design-token normalisation (Jony pass)
  • docs: add repo health audit report
  • docs: actualize README (main-vs-release note, test suite + CI report); changelog
  • ci: render backend test report as a job summary + JUnit artifact
  • chore(frontend): remove 52 unused doc-mirror TS types
  • feat(env+docs): environment matrix table; read-only-first README
  • feat(cicd): visual CI/CD flow tab (trigger -> workflows -> jobs)
  • feat(dashboard): adaptive role dashboard brief + clickable suggested questions
  • feat(role): role brief + evidence-gated suggested questions + role-shaped Ask
  • feat(risks): human-readable, review-oriented risk explanations
  • docs: reframe README as read-only project intelligence; remove MCP from public copy
  • docs: fully remove tile-outline remnants from pigeon mark (firm top/side sweep)
  • docs: clean pigeon mark — remove black border stroke + white tile, keep feet
  • docs: bigger transparent pigeon mark at README top (white tile + halo removed)
  • docs: greet visitors with the pigeon logo at the top of the README
  • 0.1.133: docs — refresh install + welcome screenshots with new pigeon logo
  • chore: sync Cargo.lock to 0.1.132
  • 0.1.132: docs — refreshed, redacted README screenshots; gitignore raw screenshot drop
  • 0.1.131: pigeon avatar in Ask + caret toggle for sources (drop 'Show')
  • 0.1.130: fix empty groups on launch + duplicate group on drag (load after backend ready, fresh-list dedup)
  • 0.1.129: declutter Ask — quiet 'Answer style' link, smaller sources bar
  • 0.1.128: consistency pass — unified eyebrows/chips/cards, single accent token, calmer Ask empty state
  • 0.1.127: calmer Home — clear lead line + Ask path, secondary cards collapsed by default
  • docs: refresh screenshot capture guide for pigeon icon + new tabs (Security/Map, file inspector, command palette, groups)
  • 0.1.126: docs — refresh README with groups, command palette, file inspector, coupling, security lens
  • 0.1.125: clickable files in Intelligence 'where to start' + honest empty state for sparse projects
  • 0.1.124: label the handbook clearly (auto-fed into answers + 'In use' badge)
  • 0.1.123: make file inspector discoverable (clickable start-here + hotspots) + clarify role-applies-instantly
  • 0.1.122: security posture lens (CI scanners + security-relevant findings)
  • 0.1.121: file inspector (ownership, coupling, blast radius, risks) + palette file search
  • 0.1.120: surface change counts in 'what changed since last time' digest
  • 0.1.119: command palette (Cmd/Ctrl-K) for repos, groups, sections
  • 0.1.118: group view full-width + default branch in group repo cards
  • 0.1.117: render markdown in group Ask answers
  • 0.1.116: group Intelligence compares (matrix/shared-vs-unique/grouped risks) + Ask polish
  • 0.1.115: drag-to-group — auto rename on create, dedup, no-op on existing membership
  • 0.1.114: change coupling + role-aware git activity card
  • Remove 'When the work happens' weekday rhythm (low signal); keep branch-flow facts
  • 0.1.113: richer git activity card — weekday rhythm + branch flow
  • 0.1.112: Map blast radius, group fixes & DnD, icon Manage menu, editable memory
  • Rating nudges: log 👍/👎 + deterministic suggestions (bigger model / rebuild context) in Ask
  • Answer feedback (👍/👎 → memory) across Ask/Investigate/groups; calmer memory card; honest Home (default branch, momentum trend, config labels)
  • 0.1.111: repo groups, Intelligence accuracy + readability, pigeon brand, map filter
  • 0.1.111: repo groups, Intelligence readability + human cloud names, project memory simplification, pigeon brand
  • Brand: pigeon app icon, in-app logo, banners + social preview; group-level memory & handbook
  • Brand: friendly messenger-pigeon app icon + in-app logo (light/dark sets); group-level memory + handbook
  • Group Ask parity (streaming, context/memory, reasoning) + group UX polish
  • Project groups: treat several repositories as one project (Ask/Home/Intelligence) — 0.1.111
  • Intelligence: merge quick Ask + Investigate into one 'Ask the map' panel
  • Intelligence tab: human-readable overview metrics, section hints, risk summary
  • Project activity: redesign into a readable briefing
  • Home polish: robustly salvage summary from truncated JSON; remove confusing weekday chart
  • Home polish: fix loading summary, meaningful file columns (modules/pipelines/config purpose), richer Sources, plain-language Project activity lead
  • Home slim pass: collapse git-activity + AI deep-analysis into expandable sections
  • Home restructure: Project Intelligence as its own tab; fix Watcher digest wall + raw-JSON deep analysis
  • Project memory: visible 'used N memory + M facts' indicator in Ask and Investigator; version 0.1.110
  • Project memory layer (Ask + Investigator) + fix: don't pass project_memory_section to general-chat prompt; version 0.1.110
  • Project memory layer: memory items + handbook + compose_project_context injected into Ask and the Investigator (recall_memory + auto-capture); version 0.1.110
  • Docs: Investigator knowledge areas + example questions; clarify one-agent-many-tools design
  • Investigator git_history tool (file ownership + recent commits) + README agents section; version 0.1.108
  • Investigator (read-only agent): ReAct loop over read-only tools, transparent trace + sources, Investigate tab; version 0.1.107
  • Project Watcher (first agent): deterministic snapshot diff + digest, watch endpoints + persisted digest, Home card; version 0.1.106
  • Release 0.1.105 — Project Intelligence + Git intelligence
  • Git intelligence: 12-week activity sparkline, weekday distribution, who-commits with share/recency, recent-commits feed, PR-merge share; version 0.1.105
  • PI M5: Cloud + References tabs, connected Map; fixes — terragrunt detection, prd/prod env tokens, infra metadata, GH Actions jobs, overview prompt; version 0.1.104 + CHANGELOG
  • M4: Python application analyzer (framework/modules/imports/deps) into the graph + Map, branch-strategy inference in Git insights, version 0.1.103 + CHANGELOG
  • Project Intelligence M3: deployment-flow rail + gaps, environment comparison, ask-the-graph (LLM constrained to facts), version 0.1.102 + CHANGELOG
  • Project Intelligence M2: Kubernetes + Helm analyzers feeding the graph, interactive Map view, role-neutral nodes/edges API, version 0.1.101 + CHANGELOG
  • Project Intelligence M1 complete: Home block (role lens, honest states, collapsed evidence) + client/types + version 0.1.100 + CHANGELOG
  • Project Intelligence M1 backend complete: HTTP routes + DI wiring + LLM overview prompt + API inventory (build/get/overview-text)
  • Project Intelligence M1 backend: graph domain + role lenses + analyzer normalizers + sqlite snapshot repo + build use case + role-lensed presenter (10 tests green)
  • Project Intelligence M1: role-neutral project graph domain + role lenses + deterministic analyzer normalizers (5 tests green)
  • Frontend: extract useSidebarCollapsed hook; App.tsx slimmer; typecheck green
  • Frontend: extract useWorkspaceJobs hook + shared errorMessage util (App.tsx slimmer); typecheck green
  • Frontend: extract usePreferences hook into src/preferences.ts (App.tsx -182 lines); README + supply-chain + startup diagnostics
  • README actualization + supply-chain (SHA256SUMS, SBOM, test report) + structured startup diagnostics instead of silent except
  • README actualization (offline wording, cross-platform install, status); release: SHA256SUMS + SBOM + test report
  • Repo polish: latest-release/downloads badges + download CTA, Troubleshooting/FAQ, roadmap link in README; issue-template config.yml; social preview cover
  • Release 0.1.99: clean Windows release + Node 24 + grouped downloads
  • Release notes: add platform-grouped Downloads section (macOS/Windows) built from uploaded assets
  • Release 0.1.97: Windows support + bump GitHub Actions to Node 24 (v6)
  • Release 0.1.97: Windows support (NSIS installer, both engines); version bump + changelog
  • Windows: build release as GUI subsystem (no console window)
  • Windows: resolve bundled backend manifest + llama-server via resources/up/up path (+ recursive fallback)
  • Windows: find backend manifest + llama-server by recursive search of the install tree (Tauri up layout)
  • Windows: launch frozen backend without a console window (CREATE_NO_WINDOW)
  • Windows build: detect .exe backend name explicitly (MSYS -x magic hid the suffix)
  • CI: temporary manual Windows-only build test (artifact, no release)
  • Windows: skip uvloop (Unix-only) via platform marker so the frozen backend builds
  • Windows support groundwork: cross-platform backend process mgmt, win llama-server fetch, NSIS release job, native folder picker
  • Release 0.1.96: README polish + custom-model docs (llama.cpp), version bump
  • Scrub internal project name from docs, comments, and test fixtures
  • README: 8-step install + first-run visual walkthrough; rename screenshots to clean names
  • Ask: offline syntax highlighting (highlight.js) for code blocks; changelog
  • Ask: offline syntax highlighting (highlight.js) for code blocks; changelog
  • Ask UI: fix code block overlapping text; distinct code-block + inline-code styling
  • Ask: complete conversation memory (native chat-message history + context-aware retrieval); fix fake providers' generate signature
  • Ask: complete conversation memory — native chat-message history (llama.cpp) + context-aware retrieval for follow-ups
  • Ask: multi-turn conversation memory via native chat messages (llama.cpp), text fallback for Ollama
  • Ask conversation memory (recent turns in prompt); llama.cpp reranker in Models; port-bind/embedding 500 fixes
  • Fix llama.cpp port-bind 500: reap orphaned llama-server before start; embedding batch-size + retry hardening
  • Reranker in Models (llama.cpp only) + toggle polish; retry transient llama.cpp embedding 500s during indexing
  • RAG: reranker foundation (port + llama.cpp adapter + Ask precision pass, off by default)
  • RAG: path/env boost + per-file diversity in hybrid search; README: local engines + retrieval sections
  • Hybrid search (BM25 + vector RRF, path-indexed); fix llama.cpp embedding 500 (pooling); consistent context window; fix stuck Scan button
  • docs: comprehensive CHANGELOG (0.1.95); release notes from CHANGELOG + commit list
  • Ask: remove verbose verification, redesign metrics, add context-window usage; fix embedding false warning; open-url command; center setup; back-to-engine
  • Center setup screens under zoom; add back-to-engine step; minimal context block; fix Get Ollama contrast (unlayered)
  • Add About section (author, GitHub, Apache-2.0); fix ruff import order; drop deleted checkpoint doc from release audit
  • Fix ruff import order; drop deleted checkpoint doc from release audit
  • Switch backend to PyInstaller onedir for fast startup; settings & UI polish
  • feat(settings): text size + default reasoning/streaming; send think flag only when reasoning is on
  • feat(settings): text size (replaces density) + default reasoning/streaming toggles, applied to all users
  • chore(cleanup): remove dead doc/packaging endpoint subsystem + dead types; trim API inventory and tests
  • chore(cleanup): remove dead backend symbols + 50 unused client.ts functions and 44 orphaned type imports
  • ui(llamacpp): split engine models into Answer models / Search model with a divider
  • fix(compare): never list embedding models — exclude catalog + custom embedders (name heuristic) from answer-model options
  • fix(compare): restrict face-off model options to the project's backend (no Ollama vs llama.cpp mixing)
  • fix(llamacpp): show 'Use this model' alongside Delete in the model detail panel, so switching is always reachable
  • feat(llamacpp): show downloaded custom GGUF models in the manager (switch/delete by repo+filename); scan storage dir into catalog
  • feat(llamacpp): add model by Hugging Face repo only — app auto-resolves a good GGUF quant (like -hf), no manual filename
  • fix(setup): models step no longer depends on transient global active-runtime match — set-up projects stop falling back to setup on switch
  • fix(ask): answer normally when no project context (empty/stale index) with a non-blocking rebuild warning, instead of a hard 'no chunks' error; drop dead diagnostic constants
  • fix(setup): llama.cpp step shows its panel with an explicit Continue (no instant jump to indexing on toggle); auto-start engine but advance only on user action
  • fix(rag): default to persistent sqlite vector store + update default-store test
  • fix(rag): default dev VECTOR_STORE to persistent sqlite so the index survives backend restarts (was in-memory → 'no chunks' after restart)
  • fix(ux): make workspace readiness a stable per-project fact (scanned+indexed+models chosen), not tied to global active runtime — stops setup/ready flip-flop on project switch and premature Ask
  • fix(ux): open workspace straight into working view (reactivate+refresh); setup detects project engine; sidebar engine badge; fix model-row wrap regression + model details as own row
  • feat(llamacpp): auto-reactivate workspace backend on open; model details + delete in Models manager
  • fix(ollama): re-activate ollama backend on first-run install so embedding matches active runtime (unblocks Install & continue)
  • fix(llamacpp): auto-start engine when models are ready; drop redundant setup note
  • feat(llamacpp): custom HF GGUF models — download by repo/filename, switch engine, persist ref across restarts
  • fix(llamacpp): active only when engine running; Models tab = installed list + Add-a-model section (Qwen out of main list); polish Compare (no collapse, font fix, tidy run button)
  • feat(llamacpp): simple setup screen (recommended model + embedder only); full model picker/switcher stays in Models tab
  • feat(llamacpp): real on-disk install state in catalog + answer-model switcher (download/switch engine, persisted); Compare lists only installed models
  • fix(llamacpp): resolve false embedding runtime mismatch; hide Ollama model manager in llama.cpp mode
  • feat(backend-modes): model-decides routing, persist+restore active engine, live-provider status, llama.cpp-aware Choose&install + Stop, backend-neutral Ask errors, engine choice before Ollama gate
  • feat: RAM bar for llama.cpp (process RSS) + route assistant identity questions to general chat with real model identity; v0.1.94
  • feat(llamacpp): advance setup flow like Ollama once engine is running; v0.1.94
  • fix(llamacpp): stage all dylibs + dereference symlink aliases; surface llama-server stderr; v0.1.94
  • fix(llamacpp): robust binary resolver (verified) + binary_path in status; intel/arm release notes
  • feat(llamacpp): desktop shell passes bundled binary path to backend
  • feat(llamacpp): engine choice + downloads working; disable download btn while running
  • feat(llamacpp): switchable embeddings + no-env wiring; DYLD path for bundled binary; fix fetch script (tar.gz)
  • feat(llamacpp): switchable embeddings, active-backend endpoint, no-env wiring; fix fetch script (tar.gz)
  • feat(llamacpp): auto-launch llama-server runtime + engine status on setup (0.1.91)
  • feat(llamacpp): embedding provider + env wiring
  • feat(llamacpp): Ollama-free backend scaffold (provider, GGUF download, bundled binary)
  • fix(lint): remove dead F841 locals; dependabot: pin React 18 majors
  • style: ruff format + lint set F,I
  • chore(oss): add Code of Conduct, CHANGELOG, FUNDING; sync lock versions
  • chore(oss): add pathspec to lock, add Code of Conduct
  • fix(creativity): make Precise temperature 0 so the setting is observable
  • polish(ui): align Ask chip font, emphasize destructive reset confirm (bump 0.1.88)
  • fix(ui): button sizing, two-scope reset, untrack junk, spacing polish
  • fix(reset): 'Reset everything' also clears saved settings (bump 0.1.86)
  • feat(home): replace Confluence/Jira with grounded LLM + TODO cards
  • feat(home): git history insights + project activity (bump 0.1.84)
  • feat(home): git history insights + project activity (bump 0.1.84)
  • feat: role-aware analysis, Stop during build, version fix, UI polish
  • feat(index): respect .gitignore and exclude nested venv/node_modules from RAG
  • v0.1.82: build-context fix + README screenshots
  • fix(privacy): don't walk the project folder (scan/changes check) on cold launch — only after a user gesture, so the macOS folder-access prompt is tied to deliberate use; v0.1.81
  • feat(home): detect on-disk file changes since last scan (backend GET /scan/changes with mtime) and show a calm banner on Home — one button re-scans, rebuilds context, and re-analyzes; v0.1.80
  • v0.1.79: auto-run analysis on first home visit
  • ui(settings): drop the empty hero card for a slim page header; single-column stacked layout (no more half-empty Appearance row); v0.1.78
  • feat(onboarding): show download percentage next to each model's progress bar during install; v0.1.76
  • v0.1.75: cleaner models current-setup, download catalog, compare
  • fix(onboarding): models step requires models actually INSTALLED (not just selected) — no more skipping to index while a download runs; refactor(settings): drop redundant App-readiness panel + hero badges (settings = preferences only); ui(models): Compare opens by default; v0.1.74
  • v0.1.73: fix models-step freeze
  • refactor(ui): redesign the Advanced inspector into clean light cards (icon + lens/files + facts + 'Edit in Settings'); drop heavy badges/dl/footer; v0.1.72
  • v0.1.71: clean home, advanced moved to settings
  • refactor(home): single clean header (drop duplicate Workspace-overview header + redundant Use-it-now panel/second Ask) — Project Understanding is the home; quiet Re-analyze when model unchanged, prominent only when a new model is selected; v0.1.70
  • v0.1.69: nav overlap + sidebar auto-collapse + tooltips
  • feat(onboarding): live scan/index progress — poll the job and show X of Y files + percent, advance on completion (fixes dead-feeling Build context); fix(models): brief hold after Install so the step advances before the button re-enables (fewer repeat clicks); v0.1.68
  • v0.1.67: Phase 2a deep analysis
  • fix(profiles): register tester + business_analyst assistant profiles (fixes 'Unknown assistant profile' error on those modes); update profile list test; v0.1.66
  • v0.1.65: adaptive Project Understanding home + Tester/BA modes
  • feat(models): redesign manager — installed models lead as compact click-to-expand rows (Ollama metadata: params, quantization, context, capabilities), Answer/Search split by divider, 'Add or change a model' collapsed below; dedupe overview READY badges; v0.1.64
  • v0.1.63: declutter + type pass on Models
  • feat(ask): Save note / Copy / Create file as compact icon buttons; refactor(ui): simplify project chip to name + readiness dot (green can-chat / red not), drop dropdown that duplicated the sidebar; v0.1.62
  • v0.1.61: project switcher chip + quieter nav
  • feat(onboarding): auto-open Ask when setup completes; fix(ui): fully hide collapsed sidebar (no leftover strip); feat(ask): move Clear chat into the chats toolbar as a tidy button; v0.1.60
  • v0.1.59: fix install gate + collapsible sidebar + create screen + models cleanup
  • fix(onboarding): Install refreshes state after selecting installed models; feat(models): simplify — drop MCP + Advanced tabs, move answer creativity to Skills and model picker to Choose & install; v0.1.58
  • fix(onboarding): stop polling storm — lightweight install/jobs poll + refresh only on real transitions; gate takeover while exit prompt open; restores Scan and Forget; v0.1.57
  • docs(readme): hero + screenshots gallery + capture guide
  • fix(onboarding): select recommended models so setup advances; reset takeover per workspace; chore(repo): dependabot, CI hardening + advisory ruff lint, README badges/TOC, move brand assets to assets/; v0.1.56
  • feat(onboarding): full-window setup takeover + inline model download progress and status checklist; v0.1.55
  • feat: real working light theme + version 0.1.54
  • onboarding: adaptive getting-ready screen + inline install; cargo sync; stronger compact; v0.1.53
  • fix(settings): compact density tightens spacing; first-run uses cat app icon; v0.1.52
  • feat(settings): skills dropdown editor (select to edit built-in or create custom); feat(ui): calm first-run welcome screen; v0.1.51
  • refactor(settings): one unified 'Skills' editor — edit all built-in skills + create your own in the same place (drop separate panel + template-scoped UI); v0.1.50
  • feat(skills): create your own custom skills in Settings (add/edit/remove, persisted) and pick them per-question in Ask's Style dropdown; v0.1.49
  • feat(ask): per-question 'Answer style' picker in dev mode — override the skill profile for the next question (project default + 5 presets); v0.1.48
  • fix(icons): faithful macOS safe-area regen (symmetric, no halo) incl sidebar logo; fix(home): drop duplicated setup-steps grid (keep live progress); code scrolls in-bubble; drop sidebar subtitle; v0.1.47
  • fix: app icons safe-area + sidebar logo; code scrolls in-bubble; drop sidebar subtitle; v0.1.47
  • fix(ask): long code scrolls inside bubble; v0.1.47
  • ask: scroll fix, slim file chips, match label, examples on empty chat; v0.1.46
  • feat(ask): redesign answer context — attached files shown as a distinct group, project sources with thin match bars (Jony-style); v0.1.45
  • ask: auto-scroll, file chips, files-in-chat panel; v0.1.44
  • feat(downloads): global active-downloads indicator in sidebar (persists across tabs, polls backend) with per-job Stop; also includes RAM indicator; v0.1.43
  • ask: live model-memory indicator during generation; v0.1.42
  • feat(ui): icons for nav (icon+label), refresh and new-project become icon buttons with tooltips; compact sidebar; v0.1.41
  • fix(settings): developer-mode checkbox no longer stretched (exclude checkbox/radio from full-width input rule); declutter Models: drop Product goal, Download history, Manual install; guided setup as dropdowns with compact meta + shorter copy; v0.1.40
  • models: one installed list grouped Answer/Search with divider; drop duplicate; v0.1.39
  • feat(models+settings): show answer & search models together with clear download intent; embedding installable in guided setup; add Settings > Reset app data; v0.1.38
  • onboarding: wizard order + lock index until search model ready; v0.1.37
  • build: robust rustup check + clearer Intel cross-build install hints
  • build: per-arch macOS DMGs (arm64 + x86_64 via Rosetta) in scripts/build_macos_dmgs.sh; fix backend packaging-python precedence; compare one-path refactor; v0.1.36
  • compare: one clear path, planning noise behind optional preview; v0.1.36
  • feat(compare): file attach + 'Use this model for this project' winner action; honest, playful copy; collapse rating to optional memory; fix glued text; v0.1.35
  • compare: file attach + polish + honest rating copy; v0.1.35
  • fix(tauri): disable native dragDrop so files can be dropped in webview; feat: smart attachment search (relevant excerpts, not whole file); file chips send as attached_documents; v0.1.34
  • feat(ask): removable file chips; fold file content at send; files-only asks; v0.1.33
  • feat(ask): refine bottom controls into switch toggles + grouped pill; drag-and-drop file analysis into composer; v0.1.32
  • feat(ask): streaming answers (SSE) behind dev mode; fix Reasoning toggle breaking non-thinking models; doc stream route; v0.1.31
  • fix(models): focused centered single-column layout; card the next-action callout
  • feat: delete-installed-model refreshes guided setup; reasoning on/off toggle in Ask (dev mode)
  • Move download progress under the button; lead Models with Current setup; bump 0.1.29
  • Download progress bar, thinking elapsed feedback, drop useless Models hero
  • Fix build: remove out-of-scope refreshInstallStatus call; bump 0.1.27
  • Capture Ollama reasoning 'thinking' field so Ask can show it
  • Use & install: show download progress and surface failures inline; bump 0.1.25
  • Show model reasoning in Ask; reorganize the setup-steps panel
  • Ollama detection + real model download progress; fix glued setup-steps text; bump 0.1.23
  • Auto-refresh Home after scan/index; fix installed-model cards; trim downloads notes
  • Fix Ask composer layout, add a model action on Models overview; bump 0.1.21
  • Clean up the Capabilities page
  • Inline scan/build confirm in Use it now; clearer model step label; bump 0.1.19
  • Single green primary on Home: make the AI setup action secondary
  • Tame heading scale, auto-refresh after jobs, trim model manager text; bump 0.1.17
  • Auto-dismiss the restored-workspace banner instead of a Dismiss button
  • Replace developer-facing model setup notes with one calm line; bump 0.1.15
  • Restore scan/index confirmation: Use it now triggers the guided flow
  • Simplify the Home Advanced section (skills + files); bump 0.1.13
  • Clearer model picker selection, leaner AI setup panel
  • Use it now is the single action hub; strip duplicate Scan from setup steps; bump 0.1.11
  • Fix selection affordance, scan progress, model manager visibility, real runtime
  • Redesign create form, de-duplicate overview, add model disk manager; bump to 0.1.9
  • Auto-select a downloaded model and label the demo model honestly
  • Prompt about leftover temporary projects on launch
  • Include persistence in bootstrap workspace response
  • Bump version to 0.1.8
  • Add permanent vs temporary projects with quit-time cleanup
  • Bump version to 0.1.7
  • Add per-project storage size, clear-index and delete-project actions
  • Bump version to 0.1.6
  • Fix packaged-app layout regressions and app icon white edges
  • Pin backend dependencies via requirements.lock and install it in all CI workflows
  • Verify workspaces/overview route precedence behaviorally instead of via app.routes internals
  • Make API inventory route test robust to newer Starlette (skip non-path router entries)
  • Pin Ollama host and model catalog in test conftest so dev shell/.env cannot leak into tests
  • Do not load backend/.env under pytest so tests run on clean defaults
  • Bump version to 0.1.5
  • Fix classless white action button and apply a consistent typography scale
  • Put generated layouts in a low-priority @layer so hand-written fixes always win
  • Order generated layouts before hand-written component fixes; repair MCP setup
  • Fix Models section nav oval and concatenated card headings
  • Fix Models section nav, oversized CTA buttons, and Settings form layouts
  • Repair component internal layouts and button roles after legacy removal
  • Remove styles.legacy.css entirely; single clean stylesheet remains
  • Cut legacy rules for shared component classes now owned by the clean layer (batch 2)
  • Add clean shared component layer (panels, headings, badges, cards, states)
  • Move legacy styles into a low-priority @layer so clean styles always win
  • Cut legacy rules for rewritten shell, sidebar, tabs, and button classes (batch 1)
  • Use brand-green for checkboxes and radios
  • Force brand-green accent and green primary-action buttons over legacy purple palette
  • Green primary buttons and brand accent, fix create-workspace step layout
  • Unify Copy, ghost, and toggle buttons into one calm dark style
  • Darken the model download history panel on the Models tab
  • Dark-patch legacy light cards and form controls across Models/Settings
  • Symmetric cat avatar for questions and a cleaner verification panel
  • Consolidate ignore rules into the root .gitignore (ignore .env everywhere)
  • Single-command local dev with backend pointed at the app data
  • Remove white background behind the sidebar app icon so its transparent corners do not show white
  • Use generated avatar art, fix app icon edges, close row menu on outside click, tint user messages with chosen palette
  • Style the Answer style and sources focus panel so labels and values are clearly separated
  • Rework saved chats history into a calm, bounded panel
  • Add cat/robot avatars, fix Copy button, separate answer from notes, refine history rows
  • Declutter the saved chats history into a calm, scannable list
  • Tidy Ask answer: lighter text buttons, compact question bubble, quiet status note
  • Start Ask with a fresh chat on launch instead of reopening the last conversation
  • Add developer-details toggle on the Ask screen and simplify the default answer
  • Fix Home status cards and compact the Ask empty state
  • Remove legacy Ask styling now replaced by the clean stylesheet
  • Refine Ask interface: calmer layout, lighter answer, scroll controls
  • Stage 2: clean Ask tab layout with sticky composer and scrolling messages
  • Darken remaining light cards in the workspace AI setup panel
  • Force consistent dark theme during migration and fix shell glitches
  • Begin staged UI cleanup: clean foundation for shell, sidebar and tabs
  • Pin app version to bottom-left so it stays visible while scrolling
  • Remove stray temporary virtualenv from version control
  • Answer general conversation in Ask instead of forcing project context
  • Keep update notice quiet while release repo is private
  • Show app version in sidebar and surface auto-update status
  • Ask UX: Enter to send, skill-aware examples, auto-dismiss startup banner; bump to 0.1.3
  • Build a .dmg installer in addition to the updater bundle
  • Release workflow: fix backend Python for CI, ship macOS first
  • Fix tests broken by earlier changes: generate() signature and icon.png size
  • Configure auto-updater
  • Wire Tauri updater: config, plugin dependency, startup check
  • Add free auto-update release pipeline and runbook
  • Hide the sources panel for general-knowledge answers (no project files)
  • Smart blend: let the model answer general questions, not only project files
  • Calm the Ask answer: render markdown, hide telemetry by default
  • Clarify Ask responses for non-project questions and the test model
  • Add answer-creativity (temperature) control for the LLM
  • Tidy remaining views: calmer Home, plainer Capabilities/Reports copy
  • Adopt new flat brand icon across app and desktop build
  • Consolidate type scale into a single source
  • Refine typography: system font base, font smoothing, heading tracking
  • Tidy Settings: plainer hero copy, hide density toggle by default
  • Enlarge brand icon artwork to fill the icon canvas
  • Tidy Models tab: plain-language hero, hide install manager by default
  • Add brand logo and app icons; wire into app and desktop build
  • Redesign Ask tab to a calm, chat-first layout
  • Add image attachment to Ask for vision models
  • Support image input for vision models in the Ask flow
  • Support an updatable model catalog fetched from a hosted manifest
  • Show download size and memory-fit verdict in AI picker
  • Hide fake models from picker; explain embedding reindex cost
  • Simplify AI model picker into recommended cards
  • Add Developer mode toggle and hide advanced sections by default
  • Remove two unused ceremony docs (repo guide, production readiness)
  • Decouple tests from documentation files
  • Rename task-numbered docs to clean topical names
  • Consolidate docs: remove redundant design-iteration files
  • Make qdrant-client an optional dependency
  • Remove orphaned one-off task-contract verification scripts
  • Remove development journal and status-snapshot docs
  • Stop tracking generated *.tsbuildinfo TypeScript build cache
  • Remove accidental copies of root tree from frontend/
  • Add plain-language help for writing Ask guidance (skills)
  • Highlight privacy in sidebar: files stay on this computer
  • Add GitHub and web-page-reader read-only MCP server presets
  • Use structure-aware document chunking for RAG indexing
  • Add docs index categorizing all documentation by topic
  • Add Apache-2.0 LICENSE and update README license section
  • Stop tracking IDE (.idea) and Rust target build artifacts
  • Cleanup: remove stale nested frontend/{frontend,backend} duplicate copies, untrack .DS_Store, harden .gitignore
  • Task 274: Fixing Models tab and Ollama LLM settings
  • Task 273: Daily-use stability + visual hardening.
  • Task 273: Daily-use stability + visual hardening.
  • Task 272: Product goal UX / Skills / MCP / Dark theme.
  • Task 272: Product goal UX / Skills / MCP / Dark theme.
  • Task 271: Product UX + Model Catalog + Settings/Ask cleanup.
  • Task 270: Daily-use UI cleanup
  • Task 269: Daily-use MVP polish,
  • Task 268: Fix Runtime Readiness UX: actionable model/runtime review
  • Task 267: Persistent RAG index after packaged .app restart
  • Task 266: Persistent RAG index after packaged .app restart
  • Task 266: Polish UI
  • Task 266: Finish full packaged product smoke: create project → scan → index → ask → restart persistence
  • Task 265: Fix packaged app SQLite path/bootstrap and finish macOS packaged smoke
  • Task 264: packaged app SQLite bootstrap + Tauri CORS fix.
  • Task 263: packaged app first-run UI clarity.
  • Task 262: packaged app backend startup diagnostics + full pytest stabilization.
  • Task 261: packaged .app Tauri bridge fix + npm allowScripts policy.
  • Task 260: packaged .app frontend bootstrap fix.
  • Task 259: macOS packaged app smoke milestone + check fixes.
  • Task 258: frozen backend startup diagnostics + PyInstaller hardening.
  • Task 257: Tauri packaged app build readiness + repo hygiene.
  • Task 256: Tauri dev smoke success recorded + next packaged-app gate.
  • Task 255: fix Tauri icon RGBA blocker + icon asset checks.
  • Task 254: fix Tauri Rust dependency conflict + target hygiene.
  • Task 253: fix Tauri Rust structure + npm registry hygiene.
  • Task 251: Add macOS packaged app smoke preflight
  • Task 250: Tauri backend health readiness gate.
  • Task 249: macOS frozen runtime + Tauri smoke runbook + continue message.
  • Task 248: fix v0.1 handoff Settings crash + real app-owned Tauri backend startup implementation.
  • Task 247: fix Settings v0.1 handoff crash + app-owned backend startup gate.
  • Task 246: frozen backend smoke contract.
  • Task 209: fix Settings audit crash + frozen runtime selection + frontend chunking.
  • Task 244: PyInstaller backend runtime PoC.
  • Task 243: staged backend runtime + macOS packaging wiring.
  • Task 242: desktop stack + runtime freeze contract.
  • Task 241: desktop technology decision record.
  • Task 240: Tauri supervisor static gate.
  • Task 239: Add desktop runtime preflight
  • Task 238: Add desktop runtime readiness handoff
  • Task 237: Prepare v0.1 source release
  • Task 236: Add v0.1 UI smoke-check handoff
  • Task 235: Add v0.1 release gate and roadmap lock
  • Task 234: Harden source release hygiene audit
  • Task 233: Prepare v0.1 source release candidate
  • Task 232: Clarify final source status and v1 roadmap
  • Task 231: Finalize GitHub repository publication files
  • Task 230: Stabilize GitHub source release
  • Task 229: Prepare GitHub source release and completion roadmap
  • Task 228: Add model context build action
  • Task 227: Clarify selected embedding and context indexing state
  • Task 226: fix Models crash + Ollama recommendations flow.
  • Task 225: Fix models screen rendering and clarify model usage flow
  • Task 224: Polish product quality and GitHub readiness
  • Task 223: Add v0.1 demo and GitHub handoff
  • Task 223: Add v0.1 demo and GitHub handoff
  • Task 222: Add release candidate audit
  • Task 221: Add Windows packaging foundation
  • Task 220: Add Tauri supervisor bridge foundation
  • Task 219: Add Tauri desktop shell scaffold
  • Task 218: Add macOS backend runtime bundle readiness
  • Task 217: Wire macOS app foundation to supervisor contract
  • Task 216: Add desktop supervisor lifecycle contract
  • Task 215: Add macOS app package foundation
  • Task 214: Lock real desktop packaging design
  • Task 213: Unify agent and MCP readiness UX
  • Task 212: Polish MCP setup UX
  • Task 211: Harden local model manager UX
  • Task 210: Add model download history and safe cancel semantics
  • Task 209: Run model downloads as background jobs
  • Task 208: Polish model download job UX
  • Task 207: Add model download job status foundation
  • Task 206: Add approved local model download execution foundation
  • Task 205: Add read-only installed model detection
  • Task 204: Design controlled backend model download worker
  • Task 203: Add safe model install draft flow
  • Task 202: Add safe local model install guide
  • Task 201: Refine calm Apple-style release candidate UX
  • Task 200: Polish release candidate UI and UX
  • Task 199: Clarify post-launch UX and desktop app target
  • Task 198: Clarify packaging roadmap and desktop app direction
  • Task 197: Add macOS desktop shortcut guide
  • Task 196: Polish desktop UI consistency
  • Task 195: Add first launch packaging readiness UX
  • Task 194: Add macOS launcher foundation
  • Task 193: Add guided local model setup
  • Task 192: add agent evidence and execution readiness
  • Task 191: add agent approval-gated execution plans
  • Task 190: add workspace MCP configs and approval gates
  • Task 189: add MCP registry foundation for safe agents
  • Task 188: add manual agent workflow tracking
  • Task 187: add safe agent capability awareness
  • Task 186: fix dark theme visual consistency and alignment
  • Task 185: finalize visual consistency and demo mode
  • Task 184: polish dark and light theme visual consistency
  • Task 183: polish Apple-style light and dark UI
  • Task 182: add branding presets and report share kit
  • Task 181: add production readiness checklist
  • Task 180: add desktop startup and last workspace restore
  • Task 179: harden generated update workflow
  • Task 178: add runtime troubleshooting assistant
  • Task 177: add backup restore and migration safety workflow
  • Task 176: add startup checklist and safer local scripts
  • Task 174: add report quality checks and source coverage
  • Task 173: add report section editor and documentation exports
  • Task 172: add custom report builder from saved context
  • Task 171: add saved report history and exports
  • Task 170: add project report generation foundation
  • Task 169: add conversation knowledge base workflow
  • Task 168: restore conversation sources and enrich exports
  • Task 167: add conversation export and reusable answer notes
  • Task 166: add conversation search pin and archive flow
  • Task 165: polish conversation details and answer history
  • Task 164: add persistent workspace conversations
  • Task 163: polish skill profile reset and diff flow
  • Task 162: add skill profile audit trail
  • Task 161: add safe skill profile templates
  • Task 160: improve skill profile visibility in Ask
  • Task 159: persist workspace skill profiles
  • Task 158: improve job details and file rules explainability
  • Task 157: clarify indexing rules preview and confirmation flow
  • Task 156: persist workspace indexing rules
  • Task 155: show job history and applied file rules
  • Task 154: add job progress and cancellable workspace jobs
  • Task 153: add LLM usage metrics and improve workspace jobs
  • Add file selection preview before scan
  • Add backend jobs for workspace scan and indexing
  • Add frontend cancellation for long-running actions
  • Apply file selection rules during workspace scan
  • Add file selection preferences foundation
  • Compact Ask assistant focus panel
  • Polish Ask chat layout and compact sources v2
  • Polish Ask chat layout and compact sources
  • Center Ask composer in workspace conversation
  • Redesign Ask as workspace conversation
  • Polish skill settings UX and button consistency
  • Connect skill context to Ask prompts
  • Add browser-local skill library presets
  • Add workspace skills and assistant focus flow
  • Add workspace setup flow and improve CI
  • Add archived workspace restore and polish create onboarding
  • Add safe workspace archive UI
  • Add branding settings and CI workflow
  • Add workspace creation onboarding UI
  • Polish Settings and close Phase 10
  • Create webpack.yml
  • Simplify Settings backup tools
  • Create pylint.yml
  • Create python-package.yml
  • Add configurable backend URL preference
  • Add Settings export import for local preferences
  • Add Settings reset for local preferences
  • Fix remaining dark theme component surfaces
  • Fix dark theme contrast tokens
  • Add browser-local Settings preferences
  • Add Settings page foundation
  • Polish final beginner UX
  • Add guided onboarding and polish Models wording
  • Polish Models lower dashboard wording
  • Simplify optional model comparison UI
  • Make model selection editor optional
  • Split Models view into simple and advanced sections
  • Simplify Activity timeline wording
  • Rename Actions tab to Capabilities
  • Simplify Overview primary CTA
  • Simplify frontend UX wording
  • Polish final frontend UX wording
  • Add Overview product status section
  • Redesign Activity tab as native timeline
  • Simplify frontend actions tab UX
  • Add Ask source progressive disclosure
  • Redesign Ask tab as conversation interface
  • Simplify frontend models tab UX
  • Refine frontend app shell
  • Add Apple-style frontend design tokens
  • Add preferred experiment model apply flow
  • Add frontend experiment history
  • Add frontend experiment rating UI
  • Add frontend model experiment run UI
  • Add frontend model experiment planning
  • Clarify frontend LLM runtime mismatch guidance
  • Add frontend scan before index guidance
  • Clarify frontend reindex guidance
  • Add frontend reindex guidance
  • Render Ask answers as safe markdown
  • Improve RAG source path citation prompt
  • Smooth refresh after model selection update
  • Document real local AI happy path
  • Polish frontend visual layout
  • Polish frontend workspace sidebar
  • Unify frontend loading error empty states
  • Unify frontend status badges
  • Frontend Overview Model Warning Navigation
  • Add overview local AI setup warning
  • Improve frontend ask sources display
  • Add frontend ask question guidance
  • Add frontend ask session history
  • Add frontend ask tab
  • Add frontend activity timeline
  • Add frontend action details panel
  • Improve activation command readability
  • Add copyable activation commands
  • Add detailed models tab to frontend
  • Add workspace tabs to frontend
  • Add frontend dev CORS support
  • Add frontend API map documentation
  • Add workspace UI action catalog
  • Improve local AI activation commands
  • Add local AI activation guide
  • Add models summary to workspace dashboard
  • Add workspace models dashboard summary
  • Add workspace models dashboard
  • Add selected embedding indexing plan
  • Add ask with selected LLM endpoint
  • Add selected model usage plan
  • Add workspace model selection status
  • Add workspace model selection state
  • Add model recommendation explanation endpoint
  • Penalize fake models in workspace recommendations
  • Add workspace-aware model recommendations
  • Add model performance summary
  • Add manual model experiment ratings
  • Add model experiment comparison summary
  • Respect fake LLM model override
  • Add persistent model experiment runs
  • Add per-request LLM override
  • Add model experiment planning endpoint
  • Make model switching plan provider aware
  • Add model switching plan endpoint
  • Add project handoff and roadmap docs
  • Add reloadable user model catalog
  • Add user model catalog file support
  • Tighten model recommendation filtering
  • Add local model catalog and recommendations
  • Stabilize project docs and API inventory
  • Add workspace metadata update endpoint
  • Add reversible workspace archive lifecycle
  • Add workspaces overview endpoint
  • Add workspace dashboard endpoint
  • Add workspace quick start endpoint
  • Add onboarding bootstrap workspace endpoint
  • Add runtime setup guide
  • Add onboarding setup command suggestions
  • Add onboarding plan endpoint
  • Add assistant profiles and workspace recommendations
  • Add runtime provider health endpoint
  • Add workspace readiness endpoint
  • Add workspace timeline backfill
  • Add persistent workspace timeline
  • Add deterministic RAG answer quality warnings
  • Improve RAG prompt grounding
  • Add optional Ollama LLM provider
  • Improve ask diagnostics for empty context
  • Add RAG ask skeleton with fake LLM
  • Add dimension-aware Qdrant collections
  • Add optional Ollama embedding provider
  • Add optional Qdrant vector store adapter
  • Persist workspace index status
  • Add workspace indexing and context search skeleton
  • Add deterministic project overview report
  • Add command activity to workspace summary
  • Add safe local command runner
  • Add command execution policy
  • Add deterministic workspace command suggestions
  • Add command approval workflow skeleton
  • Add deterministic DevOps analysis summary
  • Add deterministic GitHub Actions analyzer
  • Add deterministic Terragrunt analyzer
  • Add deterministic GitLab CI analyzer
  • Add deterministic Terraform analyzer
  • Add workspace summary with suggested actions
  • Persist latest project scan per workspace
  • Add skill registry for project scan detection
  • Add deterministic project scanner and detected skills
  • Add SQLite workspace persistence
  • Initial clean architecture backend skeleton

Full changelog: https://github.com/tonkonozhenko-mi/ai_private_workspace/commits/v0.2.0

Contributors

layer
Assets 14
Loading