LocalPilot v1.1.15

Added

• Enhance model routing and unresponsive handling; implement unresponsive model tracking and improve fallback logic for subagents

Install

code --install-extension localpilot-1.1.15.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.14...v1.1.15

LocalPilot v1.1.14

Added

• Enhance telemetry handling and provider management; add support for transient-failure retries and improve dashboard summary filtering

Install

code --install-extension localpilot-1.1.14.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.13...v1.1.14

LocalPilot v1.1.13

Added

• Dashboard charts and extra KPIs, skills across the subagent swarm, settings tab strip and theming, winget Ollama install

Install

code --install-extension localpilot-1.1.13.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.12...v1.1.13

LocalPilot v1.1.12

Added

• Dashboard — metrics, telemetry and logs, all local. LocalPilot: Open Dashboard (status-bar menu, chat title menu, Settings → Dashboard) shows what LocalPilot did on this machine over Today / 7 / 30 / 90 days / All: agent runs (done / stopped / failed, subagents), tokens in and out with tok/s per model, model calls with retries and errors, tool calls with failure rates, inline-completion acceptance, and how many providers are up — each with a per-day bar chart. Tables break it down by model (time to first token and request time as median / p95), mode (steps, duration, tools), tool (calls, failures, cached hits, share), provider health (uptime, last error, probe latency), completions by language, isolated-worktree activity and the most recent problems. A Logs card mirrors the LocalPilot output channel with level filter, search and live Follow, and opens the day's log file. Everything is computed from daily JSONL files in ~/.localpilot/telemetry that the extension and the CLI share; prompts, file contents and tool arguments are never recorded, and nothing is sent to a third party. localpilot.telemetry.enabled turns recording off; localpilot.telemetry.retentionDays (default 30, 0 = forever) prunes old files. Export… saves everything as JSON; Delete all recorded data… clears it. The extension API exposes telemetry.summary(days) and telemetry.logs().

Install

code --install-extension localpilot-1.1.12.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.11...v1.1.12

LocalPilot v1.1.11

Fixed

• CI on Windows: isolated-worktree smoke checks. repoRoot() returned git's forward-slash (and possibly 8.3 short-name) path on Windows, which never matched the canonical workspace path; it now resolves to the OS-canonical form. The smoke test also pins core.autocrlf=false in its scratch repository so worktree checkouts compare byte-for-byte on runners where Git for Windows defaults to CRLF.

Install

code --install-extension localpilot-1.1.11.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.10...v1.1.11

LocalPilot v1.1.10

Fixed

• Providers show their connection status without being asked. The status dots on Settings → Providers stayed grey until Check all connections was clicked; the page now probes on its own when it opens, again when a provider is added or its URL changes, and when the last check is older than a minute — showing checking while it runs, then N models · 12 ms or unreachable. The button remains for an immediate re-check.
• The Workers card survives reopening a chat. A finished fan-out used to collapse into a plain Ran 2 subagents in parallel row when the transcript was rebuilt; the report is now split back into the same N Done card with one collapsed lane per worker (report or error inside), including N failed. The Agents panel and @localpilot chat participant also show the checklist title for each worker.

Install

code --install-extension localpilot-1.1.10.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.9...v1.1.10

LocalPilot v1.1.9

Added

• Model gallery — search and install anything from the Ollama library. A new Settings → Model gallery page (LocalPilot: Browse Ollama Model Gallery…, also in the LocalPilot menu) browses the whole public library on ollama.com, not just the curated catalog: search as you type, filter by Tools / Vision / Thinking / Embedding, sort by popularity / newest / name, and page through results. Choose size lists every tag of a model with its download size, context length, inputs and age, marked fits / tight / too big for this machine, latest and installed; plain sizes come first and quantisation / MLX / bf16 variants sit behind show all. Install pulls through your local Ollama server with live progress and cancel, and the model shows up under Models, in the picker and in routing when done. Pull by name accepts anything ollama pull does, including user/model:tag and hf.co/org/repo:Q4_K_M GGUF repositories; references are validated on both sides. Cloud-only models and tags are shown but marked not downloadable; the page offers Start Ollama / Install Ollama when no server is running. The Models catalog links to it.

Fixed

• No more duplicated replies. When a small model announced an action it never took ("Delegated to 2 background workers. I'll report back when they finish.") and the agent loop pushed it to act, the second answer streamed straight onto the first one ("…finish.Understood as:…"). The loop now emits a nudge event: the chat dims the retracted sentence under a retracted label, explains in one line why the model is being asked again, and the real answer starts in a fresh block — live and when the chat is reopened. A reply that re-announces a hand-off after the workers already finished is nudged to report their results instead of being treated as a new claim, and a fan-out that ran counts as work done so the real report is never flagged as a claim. "Understood as: …" is shown as a quiet kicker rather than body text.
• One kind of "running" instead of seven spinners. Tabs, the background-agents strip and each worker lane show a small breathing dot; the only spinners left are the N Working card and the status line. In-progress checklist items use a filled marker instead of spinning. While workers run, the status line summarises (2 workers running · 1 done) and each lane carries its own current step — the same text is no longer shown twice.
• Worker lanes fit a sidebar. The agent name, task and current step are laid out as a grid: in a narrow panel the task gets its own line under the agent instead of being squeezed to nothing, and checklist items run with run_todos_parallel show their title (not the first 60 characters of the prompt) on the lane.
• CLI / headless runs can create files again. write_file on a new path failed with edit.createFile is not a function outside VS Code; edit_file with an empty old_text on a file that does not exist now creates it instead of throwing ENOENT, and a non-empty old_text on a missing file gets a clear hint to use write_file.

Install

code --install-extension localpilot-1.1.9.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.8...v1.1.9

LocalPilot v1.1.8

Added

• Isolated worktrees — agents never clobber each other or your open files. Every agent chat now works in its own copy of the repository (a git worktree under ~/.localpilot/worktrees/), seeded from your checkout including uncommitted work; .localpilot/worktrees.json can list setup commands (npm ci…) to run in each new copy. The changes panel shows N files changed in isolated copy with before/after diffs, Apply to workspace (changes land as ordinary uncommitted edits, 3-way merged on conflict, conflicts listed) and Discard; the Agents panel has the same actions plus Open Agent's Worktree in New Window. localpilot.worktrees.mode = always (default) / parallel / off; outside git, agents edit directly. Worktrees of deleted chats are cleaned up.
• Best of N. Run one task on several models at once (2–8), each isolated; when the last one finishes a notification offers Compare results, listing each model with file counts and its answer, and Apply this result / Apply and discard the others. Also as LocalPilot: Best of N… in the palette.
• Multitask, Debug and Plan the way Cursor does them. The mode picker now offers Agent · Multitask · Debug · Plan · Ask. In Multitask you describe the outcome ("Coordinate parallel tasks…"); the coordinator has no file-editing tools, so it must split the work and hand it to background workers (run_todos_parallel now takes the pieces inline as tasks), shown in the chat as a N Working · Stop All card with one row per worker and its current step, then N Done. Debug reproduces first, fixes the root cause and verifies. Plan ends in a Prepared plan card — title, summary, the .plan.md file saved under .localpilot/plans/, the todos — with a Build button (⌘↩) that switches to Agent and runs the checklist in parallel. A tool call printed as an argument-only JSON block ({"tasks": […]}) is now resolved to the matching tool instead of being lost, and small models get two corrective turns per run instead of one.
• Multitask launcher redesigned for non-technical users. Split the work / Try several models cards, one row per task (Enter adds a row, pasting a list fills several), Ideas for this project computed from the workspace (errors VS Code shows, uncommitted changes, missing README/tests, TODO notes, dependencies, recent prompts) with More ideas from the model, autocomplete while typing, and a plain-language footer saying exactly what will start. Agents started together form a batch; when the last finishes, Review results lists them.

Install

code --install-extension localpilot-1.1.8.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.7...v1.1.8

LocalPilot v1.1.7

Fixed

• Move the chat with VS Code's native view mechanics; the activity bar icon never disappears

Install

code --install-extension localpilot-1.1.7.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.6...v1.1.7

LocalPilot v1.1.6

Added

• Pause, steer and stop — for agents and their subagents. A Pause button (⌥Esc) sits next to Stop while an agent runs: the agent finishes its current model call or tool batch, then holds (Pausing after this step… → Paused, ⏸ in the tab, frozen progress bar) until you Resume. Anything typed while it is paused — or while it is running, via Add to this agent on Enter — is queued inline as a queued row and appended to the conversation at the next step, so you can redirect the agent without losing its work; a message that arrives as the model finishes continues the turn instead of ending it. Pausing or stopping the main agent now pauses or stops every subagent it spawned, and each lane of a parallel fan-out has its own ⏸ / ■: a stopped subagent hands the parent a stopped by the user report and the run carries on. Stopping while paused closes the pending tool calls cleanly so the next turn's history is valid. The same controls are in the tab's right-click menu, the background-agents strip, the Agents panel (sessions and their subagents, inline), the Command Palette (Pause Agent, Resume Agent, Pause or Resume Agent, Pause All Agents, Resume All Agents) and the CLI (Ctrl+Z pauses/resumes, Ctrl+C stops, typed lines become steering messages).

Fixed

• Stopping a session from the Agents panel's inline ■ acted on the active chat instead of the clicked one.

Install

code --install-extension localpilot-1.1.6.vsix

Or Extensions → ··· → Install from VSIX… in VS Code. Documentation: https://isathish.github.io/localpilot/

Full changelog: v1.1.5...v1.1.6