Skip to main content
​
v0.14.0
📊 Excel and CSV support, and OCR is 3× faster💡 Tips: The OCR upgrade needs to download an updated model file.💡 Tips: Upgrade the CLI and the skills package to their latest versions as well.🚀 New Features:
  • Excel (.xlsx) and CSV spreadsheets are now parsed, indexed and searchable. The full format list is in the Tool Reference.
  • Faster local OCR: OCR is more than 3× faster on average.
  • The crash report window gains a “Download the latest version” button, so a crash can no longer leave you unable to upgrade.
  • A new upgrade notice card in the main window makes it easier to see what each release brings.
  • Website: the brand-new changelog pages are live, and much nicer to read.
  • CLI: search and list now support the xlsx and csv file types. See the CLI guide.
  • CLI: --remote mode gains two cloud tools, so your agent can search cloud libraries and use them directly. See Remote Use.
⚡ Improvements:
  • Better ordering between text recognition, speech transcription and semantic indexing, so queueing problems are far less likely to stall everything.
  • Improved background indexing on macOS being throttled by the system.
  • A smaller installer.
  • A better guided layout for the macOS installer window.
  • The official skills now cover xlsx and csv, so “search my spreadsheets” triggers correctly.
  • CLI: better error messages from the cloud tools, easier for an agent to work with.
  • CLI: checks whether a newer skills version is available and prompts you to upgrade.
🐛 Bug Fixes:
  • A rebuild no longer deletes the whole index directory, so index metadata can no longer point at a segment that is already gone.
  • A missing index segment found at startup is now repaired automatically.
  • Fixed scanned PDFs not being findable by filename while they wait for OCR.
  • The semantic search status in Index Settings no longer sits at a full progress bar.
  • Files that cannot be parsed are no longer retried forever: oversized, encrypted or corrupted files are marked “Not indexed · Extraction failed”.
  • Opening a large spreadsheet no longer freezes the preview window — spreadsheets beyond the rendering budget open in source view instead.
  • Cloud sync no longer gets stuck on a single oversized file.
​
v0.13.0
⚡ GPU-accelerated indexing, plus RTF and legacy .doc support💡 Tips: From this release, the minimum requirement is back down to macOS 13.4.💡 Tips: The Linux packages now target a glibc 2.35 baseline, so older distributions that previously refused to install will work.🚀 New Features:
  • RTF and legacy .doc documents are now parsed and indexed, and both produce outlines. The full format list is in the Tool Reference.
  • GPU acceleration on Windows and Linux — a large speed-up for semantic indexing.
  • A hardware acceleration switch in Index Settings: drop back to the CPU at any time if something goes wrong.
  • New crash detection and diagnostic reporting: the contents are redacted, and nothing is sent until you confirm it.
  • Text PDFs without bookmarks now get an outline generated automatically.
  • Website: the dashboard gains a cloud service quota section — check your OCR / ASR / Embedding usage and trends.
  • Website: a daily Agent Skills leaderboard is now live.
  • Website: Russian is now available.
⚡ Improvements:
  • Indexing compute is allocated dynamically by CPU core count, so semantic indexing on a large library is back to its normal speed.
  • Local semantic indexing has a reworked index structure and window size: long documents index faster, and retrieval accuracy improves slightly.
  • Better scheduling for indexing tasks — the parent task is no longer crowded out by its own children.
  • A clear warning when a watched folder goes missing.
  • The built-in chatbot renders LaTeX formulas, so formula display is more reliable.
  • The cloud OCR service handles more concurrency and runs faster, and its Markdown conversion is slightly more accurate.
🐛 Bug Fixes:
  • The progress numbers on the index panel are all computed in the backend now, so “completed” can no longer exceed the total.
  • Fixed the cloud sign-in flow hanging on Windows.
  • Fixed long documents being killed by the system during embedding.
  • Fixed the body of GBK and other legacy-encoded text files being dropped.
  • Fixed indexing stopping altogether after a transient failure.
  • Fixed unscannable files driving repeated rescans.
  • Fixed some non-PDF documents being missed by the indexer.
  • Documents whose OCR failed now get another chance.
  • Fixed an occasional crash caused by window operations.
  • Fixed an encoding-related error when embedding certain documents.
​
v0.12.1
🎙️ Cloud transcription is here, plus six new note-insight skills💡 Tips: From this release, Linkly AI requires macOS 15.0 or later. Support for older systems may be revisited later.💡 Tips: Cloud transcription is included in the Pro plan — sign in to your Linkly account and subscribe to Pro first.🚀 New Features:
  • Cloud speech-to-text (ASR) released: Pro users can move audio and video transcription to the cloud from Settings → AI — faster, more accurate, and lighter on your machine.
  • Libraries can take in notes and clips by tag: add a tag filter on a library’s detail page, and the matching notes / clips join that library automatically and stay up to date.
  • Clips support tags: you can add tags the moment you save a clip — which is also what makes filtering clips into a library by tag possible.
  • Six new note-insight skills: Default Insight, Inversion Thinking, CBT Reframing, Values Clarification, Second-Order Thinking and MBTI Reading. Give your own notes a read-through. These skills are also being added to Codex and other mainstream agents.
  • New builtin skill “Save Note”: just say “note this down” or “save this to my notes” and the conversation is captured faithfully into your local notes.
⚡ Improvements:
  • Skills support localised names — the slash panel, the context badges and the skills page all show them in your own language.
  • All nine MCP tools now declare their annotations (read-only / non-destructive): clients that grant permissions by read-write tier, ChatGPT among them, no longer put a confirmation dialog in front of every read-only lookup.
  • Notes saved through MCP now carry the real application name on their source badge (ChatGPT, Claude, Cherry Studio and so on).
  • Media transcription gains a “language unsupported” state: media in a language the current transcription engine does not cover is labelled as such, instead of being reported as a plain failure.
  • Website: the help chatbot is now embedded across the dashboard — ask it your questions directly.
  • Website: the browser extension page now offers a direct package download and manual installation instructions, so you can install it even when the Chrome Web Store is out of reach.
  • Linkly AI CLI v0.6.3: a clearer message when the desktop app is not running.
🐛 Bug Fixes:
  • macOS 14 and earlier no longer crash silently at launch; they now get a clear message that macOS 15.0 is required.
  • Fixed the source badge on an MCP-saved note falling back to external-mcp after the note was edited in the desktop app.
  • Fixed a black console window flashing up on Windows while transcribing media or configuring an MCP client.
  • Fixed finished search rows flickering along with indexing progress in the settings index panel.
  • Fixed cloud library search failing on queries containing punctuation or reserved words.
​
v0.12.0
☁️ Cloud embedding is here: Pro users can switch to the cloud💡 Tips: Cloud embedding is a Pro feature — sign in to your Linkly account first and make sure you have an active Pro subscription.🚀 New Features:
  • Cloud embedding released: Pro users can switch to the cloud embedding service — faster, better semantic search quality, and lighter on your machine.
  • Notes: added timeline navigation.
  • Notes: added a tag filter rail.
⚡ Improvements:
  • Scanned files are less likely to be OCR’d twice: a scanned PDF that already carries a hidden text layer is now accepted as-is once it passes the check, instead of queuing up for another full recognition pass.
  • Magazine-style pages (“full-page image + visible text”) are no longer mistaken for scans.
  • Fixed occasional cases where indexing stalled for a long time on lower-spec machines.
  • The high-performance indexing mode now makes better use of the machine’s resources.
  • Cloud libraries: documents of a million words or more no longer fail to index, and they no longer drag down the rest of their batch.
  • Cloud libraries: indexing is now considerably faster.
🐛 Bug Fixes:
  • Fixed custom models (OpenAI-compatible, not reporting a context window) being unable to compact a conversation manually.
  • Fixed the UI freezing for several seconds when a note carried an out-of-range date (a mistyped year, for example), and the timeline dropping the most recent day in time zones where daylight saving skips midnight (Chile and others).
  • Unified how #tags in the note body are parsed, so tags are now recognised identically across the backend, the frontend and every editing path.
​
v0.11.0
🧠 Agent Skills are here: professional workflows for your Chatbot💡 Tips: A skill is simply a folder containing a SKILL.md that describes “what to do, and when”. Linkly AI ships with three default skills (Book Notes / Deep Research / Library Audit), and you can drop third-party skill folders into ~/.agents/skills and enable them after refreshing the Skills page. Type / in the input box to bring up the list of commands and skills.💡 Tips: The CLI and linkly-ai-skills need to be upgraded to the latest version, v0.6.1, alongside this release.🚀 New Features:
  • Added Agent Skills: a new Skills entry in the Chatbot sidebar, with three built-in skills — Book Notes, Deep Research, and Library Audit — that can be enabled one by one; invoke them straight from the / panel; you can also install new skills into ~/.agents/skills.
  • The list tool grows from “notes only” to three scopes (folder / library / notes), with filtering by type, modified time and tag plus pagination. It also points out the README in a directory, so AI reads the instructions before it starts. Cloud libraries are supported too.
  • Added an update-ready badge: once a new version has finished downloading, a prompt appears in the Chatbot sidebar; confirm it to restart and complete the update (you are warned first if a conversation is still generating a reply).
  • The note card menu gains “Show in Folder”, taking you straight to the note’s original Markdown file.
  • The About page, the macOS Help menu and the tray gain Website / Community / Documentation entries.
  • Linkly AI CLI v0.6.1 released: adds the list command (three scopes — folder / library / notes — filterable by type, modified time and tag) and note-save, the CLI’s first write command; read / outline / grep accept several document IDs at once, or read them from a pipe with - (linkly search --json | jq -r '.results[].doc_id' | linkly outline -); the new linkly completions <shell> generates shell completion scripts; mcp --remote reaches cloud libraries through the cloud gateway; adds auth status / auth logout; search supports --scope / --tags, and read supports --image-text.
⚡ Improvements:
  • “Rebuild Index” is upgraded to a full re-extract and rebuild
  • The note timeline loads more automatically as you scroll near the edge, so no more clicking “Load more”; the title bar shows the total note count; the publish button now has a label and a shortcut hint; expanding the editor is animated and focuses automatically.
  • After AI writes a note through MCP, the note view updates live — no need to refresh.
  • External AI clients can report their own app name when writing a note, so the source badge on a note card no longer always reads external-mcp.
  • Credit multipliers for GPT-5.6 Terra / Luna are lowered (Terra 7x → 6x, Luna 3x → 1x), and the monthly Pro allowances rise accordingly (Terra 38 → 45, Luna 90 → 270).
  • Website: library pages gain a “Use in AI” dialog that hands you a setup prompt for that library — copy it to your own AI assistant to get connected.
  • Website: Explore categories now have their own URLs and support deep pagination.
  • Website: the Terms and Privacy Policy have been rewritten to match the product we actually ship; visitors from the EEA / UK / Switzerland are asked first whether they accept analytics and advertising cookies.
🐛 Bug Fixes:
  • Fixed searches returning nothing when filtered by library or folder (when the same directory was recorded with different letter case, a whole-library search silently returned 0 results).
  • Fixed tags being wiped when AI edited a note through MCP: the #tags in the note body are now the single source of truth, tags passed in by AI are only ever appended, and a tag is removed by deleting its #tag from the body.
  • Fixed skills failing to load when SKILL.md used YAML multi-line syntax (description: > / |).
  • Fixed occasional copy failures for notes (macOS clipboard permissions), and the complete lack of feedback when copying a draft failed.
  • Fixed the tag autocomplete popup drifting off-screen once the editor was expanded; it is now anchored to the caret.
  • Fixed the session row in the sidebar staying highlighted after switching to the Notes or Skills view.
  • Fixed the note search box showing two clear buttons, and content shifting horizontally when the scrollbar appeared.
  • Fixed duplicate tool-call markup appearing in replies from certain models (DeepSeek V4).
  • Website: fixed the Dashboard model list overflowing its hover card, and the pricing page still referencing a retired model.
  • CLI: fixed list, grep --mode count and cloud library searches reporting success when there were no results; exit-code semantics are now opt-in via --exit-code and no longer affect existing scripts.
​
v0.10.0
📝 You can now take notes in Linkly AI — minimal card notes, made for agents💡 Tips: Notes are plain Markdown files kept in your local library folder. Nothing is uploaded to the cloud, and you can open them in any other editor at any time. Once written, a note is indexed automatically, so it can be searched and cited by AI.🚀 New Features:
  • Added local Notes: a new note entry in the Chatbot sidebar (⌘/Ctrl + Shift + N takes you straight there), with a card-based timeline, tag filtering, sorting by time, ⌘F full-text search, one-click copy, and expand to full text.
  • AI can read and write your notes: two new MCP tools, note_save and list, let Claude, Cursor, and other AI assistants turn a conversation into a note, append to an existing one, or list notes by tag and pick what to read.
  • The Chatbot now supports automatic context compaction: when a long conversation approaches the model’s context limit, history is compacted automatically, so you no longer have to start over and restate the background; you can also type /compact to compact at any time.
⚡ Improvements:
  • Image references in browser-extension clips are now recognized correctly: previously, a clipped article whose title contained spaces lost its image mapping for the entire document. Fixed for both local and cloud libraries.
  • Markdown preview now renders embedded HTML, so tables recovered by OCR display properly.
  • The search index readiness shown in Settings is now accurate — it no longer reports the index as usable before it has been built.
  • Pushing to a cloud library now retries automatically when the network hiccups, instead of failing the whole batch and re-uploading from scratch.
  • The website has a new home page, plus dedicated browser extension and CLI pages.
🐛 Bug Fixes:
  • Fixed temporarily unreadable files (cloud drive not yet synced, external disk not mounted) being retried too frequently and slowing indexing down.
  • Fixed the content fingerprint being cleared during a media file refresh, which could lead to a cloud copy being deleted by mistake.
  • Fixed libraries with no folders taking part in auto-sync and failing over and over.
  • Fixed the embedding progress bar going backwards when documents were added during indexing.
  • Fixed file watching not waking the indexing pipeline in some cases, leaving changes unapplied for a long time.
  • Fixed duplicate update checks firing within a short window.
  • Fixed usage statistics still being sent after being turned off, and data being dropped when a send failed.
​
v0.9.0
🚀 Text inside scanned PDFs and images is now searchable — and readable by AI💡 Tips: Linkly AI has upgraded its text-recognition (OCR) engine. Scanned PDFs and images are now recognized page by page and restored into Markdown. Scanned PDFs already in your library will be re-parsed automatically.🚀 New Features:
  • Added text recognition (OCR) for scanned PDFs and images: powered by the new PP-OCRv6 engine, with substantially improved accuracy for Chinese / Japanese / English. Scanned files are restored into structured content with headings and tables, and can be exported as Markdown.
  • The built-in Chatbot can now reference cloud libraries: mention a linked cloud library with @ and search it just like a local one.
  • Added library titles: a library can have a real title and description, shown consistently across Settings, the Launcher, and the Chatbot.
  • The Launcher now supports sorting by time: on top of the existing filters, results can be sorted by time.
  • A desktop shortcut is now created automatically once onboarding completes (and removed on uninstall on Windows).
  • Added index size compression: after upgrading, the index is compressed automatically to reduce disk usage — shrinking the index by up to 80%.
⚡ Improvements:
  • Optimized how heavy compute tasks (OCR, speech transcription, embedding) use resources, avoiding the lag and out-of-memory issues caused by running them in parallel.
  • Improved image and PDF extraction for more stable parsing.
  • “Rebuild index” is now only allowed while background long-running tasks are idle, to avoid interference.
🐛 Bug Fixes:
  • Fixed issues where certain models (such as Kimi) could produce repeated answers, lose the previous turn’s content, or call tools twice in the Chatbot; when a conversation is interrupted, the reason is now shown clearly.
  • Fixed several cases where scanned PDFs and images were not recognized, or were recognized incompletely.
  • Improved the stability of index upgrades and database migrations, fixing a case where fields required by new features could be missing on certain upgrade paths.
​
v0.8.0
🚀 Your dusty local recordings and videos — now fully usable by AI💡 Tips: Linkly AI now supports local speech transcription. Spoken content in audio and video files can be transcribed to text and indexed, so it becomes searchable and referenceable by the Chatbot just like any document. Everything runs on-device — nothing is uploaded. Enable audio / video parsing under Settings → Indexing to get started.🚀 New Features:
  • Added local speech transcription: a built-in lightweight transcription model converts speech in audio/video into indexed text, supporting Mandarin / Cantonese / English / Japanese / Korean.
    • Audio formats: mp3, wav, m4a, flac, aac, ogg. Video formats: mp4, mov, mkv, webm.
    • Audio parsing and video parsing can be toggled independently, since transcription is resource-intensive.
  • Added a standalone Appearance settings page: adjust theme, font size, and macOS Dock icon visibility.
  • Added auto-sync for cloud libraries: once enabled, new content is checked and pushed to the cloud automatically every 5 minutes.
  • The Linkly AI Space location can now be changed: move the default space directory to a custom location (the former “Linkly AI Folder” is renamed to “Linkly AI Space”).
  • Added drag-and-drop folders: drop a folder directly onto the watch list or a library’s detail page to add it.
  • Chatbot added download replies: save an entire AI reply, or just a table within it, to your downloads folder.
  • Chatbot sidebar now shows an indexing status indicator at the bottom, so you always know indexing progress.
⚡ Improvements:
  • Chatbot conversation titles are now generated as soon as a message is received, instead of later.
  • Launcher filters now include “Audio” and “Video” document types.
  • Onboarding now uses a brand-new sample library.
  • Destructive confirmation dialogs now use a consistent, more prominent warning style to reduce accidental actions.
  • Website: added Alipay / WeChat Pay for users in mainland China.
  • The library explore page now supports one-click starring, with improved popularity ranking.
  • Improved website loading speed.
🐛 Bug Fixes:
  • Fixed an issue where frequently changing files (e.g. in cloud-sync watched folders) repeatedly interrupted indexing, preventing large libraries from ever finishing.
  • Fixed semantic search failing after the embedding model was unloaded due to idle sleep.
  • Fixed a possible crash when installing updates on macOS: updates are now installed on next restart, and a downloaded update package is preserved even if installation fails, so it doesn’t need to be re-downloaded.
  • Fixed Free users being incorrectly blocked from querying cloud libraries as if it were a Pro-only feature.
  • Fixed indexing and cloud push silently failing for some users due to missing database fields.
  • Fixed a potential 404 / 308 redirect storm on library share links.
​
v0.7.0
🚀 Linkly AI Pro is here!💡 Tips: After 4 months of iteration, Linkly AI has launched the Pro paid plan, bringing more cloud capabilities and more curated models.🚀 New Features:
  • Added official model quota: view more detailed model quota, remaining usage, and more in Chatbot and Web Dashboard.
  • Added Free / Pro model access: Free users can continue using base models, while Pro users will gradually unlock curated frontier models such as DeepSeek V4 Pro and Qwen3.7 Plus.
  • Cloud libraries now use plan-based quotas: Pro supports more cloud libraries, higher file limits, and more usage calls.
  • The desktop receiving flow for browser clips is complete; the browser extension can now save web content directly into the Linkly AI Space for indexing.
  • Chatbot sidebar now includes Settings, website, and quota entries, making common actions easier to find.
  • Added one-click generation for cloud library introductions.
⚡ Improvements:
  • Pricing page and Dashboard now explain Pro benefits in more detail, including models, cloud libraries, MCP Tunnel, and other capabilities.
  • Web Dashboard improves API Key creation, library filtering, starring, and connection status display.
  • After accepting a cloud library invitation, you can connect it directly to your own MCP service.
  • Improved indexing and settings-save stability, reducing repeated retries, state overwrites, and configuration sync issues.
🐛 Bug Fixes:
  • Fixed cases where the tray menu could crash or show out-of-sync status.
  • Fixed stability issues related to MCP configuration changes, port refresh, and local service state.
  • Fixed cases where the indexer could stop unexpectedly or repeatedly retry during delete, retry, and concurrent scan scenarios.
  • Fixed several edge cases in browser clip saving, deletion, and scanning, improving file safety and reliability.
​
v0.6.1
🚀 Publish cloud libraries publicly, plus filtering and starring🚀 New Features:
  • Cloud libraries can now be published with two visibility options, Public and Showcase, making them easier to share and showcase
  • “My Libraries” filtering: filter by source (All / Connected / Mine / Shared / Starred) to quickly find the library you need
  • Library starring: mark frequently used libraries as starred with one click for quick access
⚡ Improvements:
  • After accepting a cloud library invitation, the page now guides you to link in one click right there, with no need to navigate to the library page
  • Improved the reliability of the account sign-in flow, fixing an occasional issue where sign-in could not complete
​
v0.6.0
🚀 Document preview, EPUB e-books, and model sleep arrive; a brand-new onboarding flow and account sign-in💡 Tips:
  • Upgrading Linkly AI CLI to v0.5.1 is recommended (adds EPUB support)
  • Updating linkly-ai-skills to the latest version is recommended
  • MCP tool descriptions have been updated (added EPUB and cloud library ownership tags); your connected AI tools may need to reload the MCP server
🚀 New Features:
  • Added account sign-in: sign in to your Linkly cloud account in one click via the browser (OAuth), with no more manually entering an API key
  • Added EPUB e-book support: .epub files now join full-text and semantic search, and AI can search and read them via MCP/CLI; after upgrading, e-books in already-indexed folders are picked up automatically — no action needed
  • Added document preview: preview document content directly in the Launcher and the Chatbot workspace, with a one-click source / rendered toggle for Markdown and HTML
  • Brand-new onboarding: the first-launch onboarding screens have been rebuilt with a simpler flow, and a multilingual sample library is prepared automatically
  • Added model idle sleep: the embedding model automatically unloads after a period of inactivity to free up memory and wakes on demand; toggleable in Settings
⚡ Improvements:
  • Cloud libraries are now tagged with ownership in MCP list_libraries / explore ([yours] / [shared]), so AI can tell library sources apart more clearly
  • The embedding model is prefetched when the Launcher is shown, speeding up the first search
  • Chatbot error messages are now categorized by type (sign-in / permission / model), with a corrected fallback model selection; the available model list refreshes automatically after sign-in
🐛 Bug Fixes:
  • Fixed file watching repeatedly firing delete events and flooding logs because of cloud-sync folders (e.g., temporary files from Synology or Dropbox)
  • Fixed the MCP command-line integration sometimes failing to connect because the PATH environment variable was truncated
  • Fixed the index data size being shown inaccurately on the Settings page
  • Fixed a manual index retry in Settings not actually restarting
​
v0.5.1
📊 PowerPoint support is here; Chatbot gains @ context and a workspace💡 Tips:
  • Linkly AI CLI needs to be upgraded to v0.5.0 (older versions reject --type pptx)
  • Updating linkly-ai-skills to the latest version is recommended
  • MCP tool descriptions have been updated; your connected AI tools may need to reload the MCP server
🚀 New Features:
  • Added PowerPoint support: .pptx files now join full-text and semantic search with a purpose-built outline index, and AI can search and read them via MCP/CLI; after upgrading, PPT files in already-indexed folders are picked up automatically — no action needed
  • Chatbot gains @ context: type @ to target a library, folder, or file and scope the AI’s answer to exactly that content
  • Chatbot gains a workspace: files referenced in a session are gathered automatically, with file preview and browsing of indexed library files
  • Added a unified Cloud Services settings page to manage the credentials the client uses for LinklyAI cloud services
  • Chatbot is now the default startup window
⚡ Improvements:
  • MCP search’s path_glob now uses substring matching: @folders and full paths hit directly, making folder-scoped retrieval more accurate for Agents
  • MCP find_paths now returns a ready-to-use path_glob, making folder search easier for Agents
  • Long session titles are truncated, and the Launcher shortcut is shown in Chatbot
  • The shortcut settings page now includes the newly added shortcuts
  • Many Chatbot UI refinements (session list scrolling, source list styling, @ picker appearance, and more)
  • App update downloads now retry automatically on failure, with a friendlier error message
  • Added folder access loss detection (macOS): after an app update the system may revoke folder access, leaving new files unindexed; Linkly AI now detects this at startup and guides you to re-authorize — one confirmation per folder in a system dialog, no need to remove and re-add folders; inaccessible folders show a ⚠️ badge in Settings → Folders
🐛 Bug Fixes:
  • Fixed PDF bookmark outlines possibly being lost after rebuilding the index
  • Fixed the cloud library MCP tunnel occasionally reconnecting every two minutes
  • Fixed folder permissions on macOS causing cloud library push failures
  • Fixed out-of-memory errors when indexing large documents in cloud libraries
  • Fixed Chatbot input being lost and errors going unreported when sending a message failed
  • Fixed some file previews failing to open and empty files showing no placeholder
  • Fixed model selection issues caused by chat data from older versions (auto-repaired on startup)
  • Fixed layout issues such as the notification toast squeezing the sidebar and the divider occasionally getting stuck
  • Fixed the “Check for Updates” button becoming unresponsive after a failed update until the app was restarted
  • Fixed semantic indexing retrying endlessly and wasting resources on documents containing special characters
​
v0.5.0
☁️ Cloud Library is here — sync local libraries to the cloud in one click, online 24/7, shareable💡 Tips:
  • Before using the cloud library for the first time, sign in on the Linkly web app and set your username
  • Linkly AI CLI needs to be upgraded to the latest version
  • linkly-ai-skills needs to be upgraded to the latest version
  • MCP has significant changes; your connected AI tools may need to reload the MCP server
🚀 New Features:
  • Added Cloud Library: in Settings, bind any local library to a cloud library and push it, after which the content stays searchable from the web app and external AI tools even when the desktop client is shut down or offline; free during the public beta
  • Added a native macOS application menu: including App / Edit / Window / Help menus and more
  • The Chatbot conversation list now shows status indicators: generating and unread messages are clear at a glance
  • Chatbot table export: tables in AI replies can be exported to CSV / Markdown to a system directory with one click
  • Settings → Folders now includes “Open containing folder”
⚡ Improvements:
  • Refreshed Chatbot window look (macOS): native system frosted glass, improved title bar; the window remembers its size automatically; the sidebar collapses adaptively with the window width
  • Improved the global right-click menu so text in the chat area can be selected and copied normally
  • Improved built-in Chatbot replies: smoother, more concise responses with more readable formatting
  • Improved the Chatbot Stop button so it halts instantly when clicked
  • Improved Chatbot code block rendering for a lighter, cleaner look
  • (MCP / external AI tools) Improved the direct return value of search results, compatible with file paths from both local and cloud libraries
🐛 Bug Fixes:
  • Fixed MCP call failures caused by some large models generating non-standard parameters
  • Fixed a window remnant and drag failure when a table was in fullscreen
  • Fixed Chatbot messages failing to save due to timing issues
  • Fixed an occasional race condition during background cleanup after clicking Stop
  • Fixed lax validation of local library names
  • Fixed folder document counts not refreshing and showing stale values after scan / watch batch processing
  • Fixed transient delete events during indexing that could cause accidental deletions
​
v0.4.1
🛠️ MCP and CLI tools leveled up — retrieval gets stronger💡 Tips:
  • Linkly AI CLI needs to be upgraded
  • linkly-ai-skills needs to be upgraded
  • After upgrading, your connected AI tools may need to reload the MCP server
🚀 New Features:
  • New MCP/CLI tool find_paths: fuzzy-match keywords against real folder names; pair with the path_glob parameter of the search tool to scope retrieval to the intended directory more precisely
  • The MCP/CLI search tool gains time filtering and sorting, supporting fuzzy time-range queries like “notes from last month” or “the most recent document”
  • New “silent launch” mode for auto-start: the window no longer pops up at startup — the app stays in the background
⚡ Improvements:
  • All MCP/CLI tool responses now carry the current timestamp (_meta.now / [meta] now=), letting the LLM compute dates from it
  • Stronger input validation on MCP tools, preventing tool-call failures caused by the LLM misspelling parameters
  • The built-in Chatbot has learned to use find_paths to locate the path before searching documents
  • Improved image OCR pre-filter: switched from a file-size check to a pixel-shape classifier, fixing cases where small images couldn’t be indexed
  • Unrecognizable images are no longer retried by the Chatbot, saving time during conversations
  • The first-run startup is more tolerant of environment issues, reducing initialization-failure crashes
🐛 Bug Fixes:
  • Fixed outline silently swallowing errors on invalid pagination or non-existent doc_id
  • Fixed search misreporting DB errors as “Library not found”
  • Fixed the desktop UI showing stale “Modified at” values during the reindex window
  • Fixed an edge case where newly created or modified files were not auto-detected
  • Fixed “Pause indexing” being ignored while a scan was in progress
  • Fixed search results not refreshing promptly after a file was renamed
  • Fixed occasional search-result anomalies when switching libraries or rebuilding indexes frequently
  • Fixed Windows auto-start occasionally not taking effect
  • Fixed database optimization failing
  • Fixed find_paths returning incorrect results in cross-library queries
  • Fixed search’s date filter not including files from the end-date day itself
  • Fixed time sorting where some newer files weren’t ranked first
  • Fixed an occasional error when older MCP clients connected to a newer desktop client
​
v0.4.0
🤖 Chatbot is here — chat directly with your local files!🚀 New Features:
  • Released Linkly AI Chat, chat directly with your local files inside Linkly AI — no need to install or integrate a third-party Agent
  • Linkly AI Chat supports multiple Providers: built-in Linkly Official, plus freely configurable OpenAI, Anthropic, DeepSeek, Ollama, and more
  • Settings now includes a Data Privacy panel, giving a centralized, transparent view of where your local data flows
  • New toggle to show full file paths: hidden by default in MCP tool results and CLI output for privacy; you can now opt in to show them
  • Released the Linkly Official curated model service, with two trial models: Qwen 3.5 Flash and DeepSeek V4 Flash
  • New in-app Changelog view: read updates directly from the launcher, no need to open a browser
  • AI Settings now show local model download progress and status
  • Launcher gains a “Pin” button: when enabled, the launcher stays open even when it loses focus or you click outside
  • Desktop client now supports Japanese, French, and Spanish
  • Added MDX file support: .mdx files commonly used in blogs and technical docs can now be indexed and searched
🐛 Bug Fixes:
  • Fixed an issue where the embedding model failed to download in certain network environments, blocking initial indexing
  • Fixed a display issue where the settings page scrollbar overlapped with content
  • Fixed item selection in the launcher occasionally drifting or becoming unstable
  • Fixed several issues causing crashes and freezes on the Windows client
  • Fixed one-click Claude Desktop integration that previously failed; the app now guides users through installing Linkly CLI on first integration when it’s missing
​
v0.3.5
  • [Improved] Back up the Windows User PATH environment variable before modifying the registry, so it can be restored if anything goes wrong
  • [Bug Fix] Fixed an issue on Windows where the one-click Linkly CLI setup could clear the user’s User PATH environment variable (only v0.3.4 was affected)
​
v0.3.4
  • [New] Added Integrations settings page — configure Linkly AI for Claude Code, Cursor, Codex, Cherry Studio and other MCP clients with one click
  • [New] Added one-click Linkly CLI installer — automatically downloads the binary and configures PATH, no terminal commands needed
  • [New] Added one-click Linkly Skills download — quickly integrate Linkly search into clients that support Agent Skills
  • [New] Added NEW badges to sidebar navigation to help discover new features
  • [New] Rewrote the first-run MCP integration onboarding step and completed UI internationalization
  • [Improved] Unified icon style across Integrations and MCP-related settings
  • [Bug Fix] Fixed a resource cleanup timing issue when quitting the app on macOS
  • [Bug Fix] Fixed windows being unexpectedly repositioned when navigating between pages inside the same window
  • [Bug Fix] Fixed an issue where external links and dialogs could not open inside the Settings window
  • [Bug Fix] Fixed Codex icon display and minor sidebar layout issues
​
v0.3.3
  • [New] Support adding directories from NAS and external drives to your knowledge base (Windows and macOS only for now)
  • [New] Support upgrading to Linkly AI Pro — lock in early bird pricing and get early access to more cloud services
  • [Bug Fix] Fixed hidden windows triggering UI freeze detection
​
v0.3.2
  • [New] Filename quick indexing: after adding a directory, indexing completes in seconds — search by filename immediately without waiting for full-text indexing
  • [New] Indexing performance mode: choose between Auto, Performance, and Balanced modes
  • [New] Power saving mode: when enabled, pauses high-power tasks while running on battery
  • [New] Independent toggles for Semantic Search and OCR
  • [New] Support selecting multiple folders at once when adding directories
  • [New] Beta program: opt in via Settings > About to get early access to new features
  • [Improved] Enhanced index status page with progress and status display for all four index types
  • [Improved] Smart indexing order: prioritizes lightweight formats like plain text and Markdown
  • [Improved] Enlarged default settings window size for easier access to more settings
  • [Bug Fix] Fixed folder removal not checking library references, causing accidental deletion
  • [Bug Fix] Fixed occasional UI freeze during file indexing
  • [Bug Fix] Fixed unstable sorting in index statistics chart
​
v0.3.1
  • [Improved] Image OCR optimization: reduced OCR resource usage, skip images with little text, improved handling of oversized images
  • [Bug Fix] Fixed app freezing on Windows when indexing large collections of high-resolution images
  • [Bug Fix] Fixed inability to record Alt+Space shortcut on Windows
  • [Bug Fix] Fixed unfriendly error messages when a shortcut is already in use by another application on Windows
​
v0.3.0
📚 Libraries are here! Organize your documents into libraries for more precise search within specific domains.
This upgrade will update existing index structures. Search results may be inaccurate during the update — please wait for it to complete before continuing use.
🚀 New Features:
  • [Desktop] Library Management — Create, edit, and delete libraries in Settings; organize folders into named libraries for structured management. View guide
  • [Desktop] Library-Scoped Search — Type @ in the launcher to enter library-scoped search, pinpointing documents within a specific library; press Backspace to exit the scope
  • [Desktop] Document Type Filter — A filter panel automatically appears when search results contain multiple types; use Cmd/Ctrl+↑/↓ to quickly switch
  • [Desktop] Search Result Highlighting & Path Optimization — Keyword highlighting in titles and paths, smart path shortening, and opening files now reveals them in the file manager
  • [MCP] New list_libraries tool for listing all knowledge libraries, usable with search and other tools
  • [MCP] New explore overview tool — AI Agents can quickly understand the topic distribution, directory structure, and top keywords of a document collection or specific library
  • [MCP] search tool now supports library and path_glob filter parameters
  • [CLI] New list-libraries and explore commands
  • [CLI] search supports --library and --path-glob flags
  • [CLI] New linkly doctor diagnostic command to help AI Agents self-diagnose issues
  • [Skills] Added library support and troubleshooting guide, enabling Agents to search libraries and resolve common issues independently
  • [Skills] Added usage instructions for list-libraries and explore
  • [Web] Dashboard now supports username settings for the cloud library namespace
✨ Improvements:
  • [Desktop] Search tool now returns up to 100 results, increased from 20
  • [Desktop] Added search result caching — repeated searches for the same keywords return instantly for a smoother experience
  • [Desktop] Improved error messages with clearer descriptions
  • [Desktop] Optimized tray menu state refresh to reduce frequent refreshes causing the menu to close unexpectedly
  • [Desktop] Launcher footer shortcut hints can now be collapsed to reduce visual clutter
  • [Desktop] Improved keyword extraction quality by automatically filtering URLs, file paths, and other noise
  • [CLI] Added 30s connection timeout and 60s tool call timeout to prevent process hanging during Agent calls
  • [CLI] Error messages improved with troubleshooting advice and linkly doctor guidance on connection failure
🐛 Bug Fixes:
  • [Desktop] Fixed settings panel occasionally displaying in English
  • [Desktop] Fixed brief flash of “no results found” while typing search keywords
  • [Desktop] Fixed DOCX document headings not being correctly extracted as Markdown format
  • [Desktop] Fixed some documents (CSV, images with OCR in progress) failing to preview
  • [Desktop] Fixed missing borders on the search box and onboarding page on Windows/Linux
  • [Desktop] Fixed title and path text overflow in search results
  • [MCP] Fixed occasional request timeout or no response issues with tunnel
  • [CLI] Fixed unable to connect with certain custom port configurations
​
v0.2.4
  • [Bug Fix] Fixed Windows cold-start issue where the launcher failed to appear or froze immediately after launch
  • [Bug Fix] Fixed Windows tray icon briefly showing an old icon on startup
  • [Improved] MCP settings page copy improved with integration-related links
  • [Improved] Onboarding now requires explicit usage selection before proceeding
​
v0.2.3
  • [Improved] Significantly faster OCR image recognition: automatically skips small icons and thumbnails, ~3x faster processing for high-resolution images, reduced CPU usage
  • [Improved] Better search result quality: images with no text content no longer occupy search result slots, more accurate semantic search
  • [Bug Fix] Fixed an issue where images pending indexing might not be processed after OCR model finished loading
  • [Bug Fix] Fixed CLI failing to run on some OpenClaw host Linux systems due to GLIBC version incompatibility
​
v0.2.2
🎉 Linkly AI can now recognize text in images!🚀 New Features:
  • Local OCR support — automatically recognize text in PNG, JPG, BMP, WEBP images and build searchable index, supporting Chinese, English, Japanese, and Latin languages
  • Redesigned onboarding flow with a conversational typing effect
  • Shortcuts settings page showing all keyboard shortcuts
  • System notification on first launch after update, with an update banner in the Launcher for quick access to the changelog
  • Changelog link added to the About page
✨ Improvements:
  • Remote tunnel connection stability improved with automatic reconnection
  • Colored brand icon for system tray on Windows/Linux
  • CLI search filter now supports image document type
🐛 Bug Fixes:
  • Fixed Metal GPU assertion crash on app exit (macOS)
  • Fixed incorrect document titles in some cases
​
v0.2.1
  • [Bug Fix] Fixed a visual issue where a gray border appeared around the window on Windows
  • [Bug Fix] Fixed tunnel toggle state not syncing after the backend auto-reconnects
  • [Bug Fix] Fixed CLI using the wrong health endpoint in --remote mode
​
v0.2.0
🚀 Major update — Linkly AI is now securely accessible over the internet!🚀 New Features:
  • MCP now supports remote connections: access your local knowledge base directly from cloud AI assistants and Agents like ChatGPT, Claude, OpenClaw, and more
  • MCP now supports LAN connections: share your MCP knowledge base service within your local network for multi-device and Agent access
  • CLI adds remote mode: connect to your knowledge base from any computer via linkly auth and --remote, making it easy for Agents to call
  • New AI tool integration guides added to the docs site, including ChatGPT, Claude, Cursor, Windsurf, GitHub Copilot, OpenClaw, n8n, Raycast, and more
  • Web Dashboard adds tunnel creation and remote call statistics panel
✨ Improvements:
  • Shimmer animation added to the indexing progress bar
  • Indexing progress now shows separately by search type (keyword / natural language)
  • Skip embedding model warm-up when vector search is disabled, speeding up startup
  • MCP settings area redesigned with collapsible card layout
🐛 Bug Fixes:
  • Fixed cache consistency and space not being released after document deletion
  • Fixed ghost records of deleted documents appearing in search results
  • Fixed long loading times when reading documents with missing cache (now silently repairs in background, available on next search)
  • Fixed old indexing tasks not being interrupted when switching watch folders
  • Fixed settings page content area layout shift
  • Fixed Windows shortcut key incorrectly using Win key instead of Ctrl key
  • Fixed “VCOMP140.DLL not found” error on Windows after installation — the installer now automatically installs the VC++ runtime dependency
​
v0.1.11
  • [New] Search results auto-deduplication: duplicate results with the same filename and size are merged for cleaner results
  • [Improved] Faster indexing with reduced resource usage during document processing
  • [Improved] Simplified tray menu indexing status display with a new “View Index” shortcut
​
v0.1.10
  • [New] Added grep tool to MCP and CLI — locate specific lines in documents by regex pattern. AI Agents can now pinpoint target content without reading page by page, significantly reducing tool calls and context consumption
  • [New] CLI now available via Homebrew and crates.io
​
v0.1.9
  • [New] Added index management page with index statistics and one-click rebuild
  • [New] Added CLI and Skills quick links to MCP settings page
  • [New] Added anonymous usage data collection with opt-in/opt-out toggle in settings (privacy-first, requires user consent)
  • [Improved] Unified indexing progress display across tray and settings page
  • [Bug Fix] Fixed index and extracted cache not being cleaned up after document deletion
  • [Bug Fix] Fixed PDF indexing failure on Windows/Linux
  • [Bug Fix] Fixed Launcher window not showing on Windows startup
  • [Bug Fix] Fixed inconsistent keyboard shortcuts across platforms
  • [Bug Fix] Fixed incorrect x86_64 download link on the website
  • [Bug Fix] Fixed broken links in website header and footer
​
v0.1.8
  • [New] Added macOS Intel (x64) build — Intel Mac users can now download and use Linkly AI
  • [Bug Fix] Fixed an issue where the index pipeline was not triggered after file watcher detected changes, and improved directory deletion handling
​
v0.1.7
  • [New] Released Linkly AI Skills — AI assistants can now better leverage Linkly AI’s tool workflow
  • [New] Released Linkly AI CLI tool — AI assistants can directly call the CLI for document search and reading
  • [New] CLI supports automatic discovery and connection to the app
  • [Improved] MCP tools now support structured output format
  • [Bug Fix] Fixed an issue where dragging the window on Windows would unexpectedly close it
​
v0.1.6
  • [Bug Fix] Fixed transparent window and window dragging compatibility issues on Windows
​
v0.1.5
  • [New] New 4-step setup wizard with permission checks for a smoother first-use experience
  • [New] Added copy file path keyboard shortcut
  • [New] Support for controlling macOS Dock icon visibility
  • [Improved] MCP outline tool now supports tree-format output, enabling AI to better understand document structure
  • [Improved] Improved PDF outline extraction accuracy
  • [Improved] Tray menu now supports multiple languages
  • [Improved] Optimized indexing performance with reduced resource usage
​
v0.1.4
  • [Improved] Optimized MCP tool descriptions for more accurate AI invocations
  • [Improved] Search result scores now displayed as percentages for better clarity
  • [Bug Fix] Fixed title misdetection for CJK (Chinese, Japanese, Korean) text
  • [Bug Fix] Improved window positioning in multi-monitor setups
​
v0.1.3
  • [New] Added MCP server — AI assistants can now directly search and read your local documents
  • [New] Support for hybrid keyword and semantic search for more comprehensive results
  • [New] Added document outline extraction
  • [New] Added multi-language support
  • [New] Support for custom global shortcuts
  • [Improved] Upgraded semantic search engine for better search quality
  • [Improved] Improved Chinese word segmentation and keyword extraction quality
​
v0.1.2
  • [New] Added full-text search engine with support for searching body content of PDF, Word, and other documents
  • [New] Launcher integrated with search, with result preview support
  • [Improved] Redesigned search result layout
  • [Improved] Window theme now automatically follows app settings
​
v0.1.1
The first usable version of Linkly AI Desktop, supporting macOS, Windows, and Linux.
  • [New] Automatic scanning of specified directories to build file indexes with incremental sync
  • [New] Brand new launcher interface with global shortcut for quick access
  • [New] Built-in auto-update with new version notifications via the tray menu
  • [New] Support for general settings and folder management