0.146.2

• Fix 400 on GitHub Copilot Claude adaptive-thinking models (e.g. Opus 4.8) when no variant is selected, by defaulting to adaptive thinking. #528

0.146.1

• Wait for rate-limit reset from response headers (Anthropic, OpenAI-compatible) on 429, showing reset time in chat; optional provider config rateLimitMaxWaitSeconds caps the wait.

0.146.0

• Discover GitHub Copilot model APIs and reasoning variants from its authenticated /models endpoint.
• New chat.autoSyncSystemPrompt config (default false): existing chats keep their system prompt and LLM prompt cache stable, changes apply to new chats. New /sync-system-prompt command to apply them to the current chat. Skills changes never auto-sync.

0.145.1

• Notify in chat when instructions change mid-chat (prompt, contexts, rules, skills, tools), invalidating the prompt cache.

0.145.0

• Add built-in variants for gpt-5.6 models (gpt-5.6-luna, gpt-5.6-terra, gpt-5.6-sol): none, low, medium, high, xhigh, max.
• Support provider-level extraHeaders, sent on completion and models list requests. #517
• Improve /init prompt, AGENTS.md, and development docs.
• Improve chat/queryContext/chat/queryFiles: rank results (basename matches first) keeping a stable order, include directories matching the query, list directory contents for workspace-relative queries like src/foo/, and refresh the workspace file index every few seconds so new files show up in completion.

0.144.0

• Add built-in variants for glm-5.2 (case-insensitive): none, medium, high.
• New /export and /import commands to transfer a chat between machines as a structured, resumable file preserving model, variant, usage and chat-id. #28
• Add Claude Sonnet 5 support.

0.143.1

• MCP: detect a broken/logged-out session (expired/revoked token, expired session) and re-initialize instead of hanging the tool call until timeout; relies on plumcp 0.2.2 correlating HTTP error responses to the originating request.
• MCP OAuth: proactively refresh an expired token before a tool call instead of waiting for the server to reject it.
• Bugfix: remote HTTP endpoints (/chats, /session, /chats/:id) now use the chat map key as the authoritative id instead of the record's :id field.

0.142.2

• /context now shows where auto-compaction triggers: a 🔲 marker on the threshold cell of the grid plus an Auto-compaction at N% line.
• Include autoCompactPercentage in the usage chat content so clients can mark the auto-compaction threshold on their context-usage bar.
• Bugfix: kill-job! commits :killed before destroying the process, fixing a race where the exit monitor overwrote it with :failed (exit 143). (#510)
• Bugfix: enabling trust mode while a subagent is running now applies to it; subagents (including nested) inherit the parent chat's trust. (#504)

0.142.1

• Bugfix: postRequest hook no longer fires for subagent chats; subagents use subagentPostRequest only, preventing duplicate hook notifications. (#505)
• Bugfix: clearer rejected tool-call result wording so models no longer assume a rejected edit was applied; it now states the call did not run and changed nothing. (#507)
• Bugfix: ask_user normalizes the options argument so a malformed value (e.g. a stringified options an LLM sometimes emits) no longer reaches clients as broken choices. Accepts string/object arrays, recovers a JSON-encoded string, and drops unusable input.

0.142.0

• Add /context command showing context-window usage with a fixed 10x10 proportional emoji grid to the left of a two-section legend: Instructions (system prompt, rules, skills, AGENTS.md, tool definitions) and Chat (tool calls, conversation), plus free space. Each category has a canonical distinct color/emoji swatch.
• Include contextBreakdown in the usage chat content so clients can render a live context-usage bar. Each category carries a canonical color and emoji (plus freeColor/freeEmoji for the unused portion) so all clients render the same colors/legend.