https://dommango.github.io/agentmats/kimi-code/ 2026-09-19T09:00:00Z https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-2-0-2 2026-09-19T09:00:00Z <ul><li><b>FIX</b> <code>/resume</code> — new messages no longer land at an old position in the conversation after resuming a session</li><li><b>FIX</b> <code>/compact</code> — no longer fails after switching to a model with a smaller context window</li><li><b>CHG</b> the agent no longer assumes the current working directory is the project root</li><li><b>FIX</b> a message sent while the agent was running no longer sometimes appears twice in the chat</li></ul> https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-2-0-1 2026-09-18T09:00:00Z <ul><li><b>ADD</b> <code>[watch]</code> / <code>KIMI_CODE_WATCH</code> — disable filesystem watchers for <code>local.toml</code>, <code>AGENTS.md</code>, skills and MCP config entirely</li><li><b>CHG</b> <code>[providers."&lt;id&gt;"]</code> — <code>api_key_env</code> names an environment variable to read the API key from, as an alternative to <code>api_key</code></li><li><b>ADD</b> <code>[database]</code> — <code>searchSyncSessionCap</code> / <code>searchSyncDebounceMs</code> tune the incremental background search-sync pass (marked <code>unverified</code> — not yet in the config reference)</li><li><b>CHG</b> <code>/yolo</code> / <code>-y, --yolo</code> — Ask When Needed no longer prompts for a <code>Bash</code> command that cannot be statically analyzed</li><li><b>CHG</b> <code>kimi install-app</code> — renamed to <code>kimi install-desktop</code>; the old name keeps working as a hidden alias</li></ul> https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-2-0-0 2026-09-17T09:00:00Z <ul><li><b>ADD</b> <code>/desktop</code> / <code>/install-desktop</code> — opens the Kimi Code desktop app page in your browser</li><li><b>ADD</b> <code>kimi install-app</code> — prints and opens the desktop app page (renamed to <code>kimi install-desktop</code> in 2.0.1)</li><li><b>ADD</b> <code>[markdown].mermaid</code> — renders fenced mermaid code blocks in the terminal as diagrams (marked <code>unverified</code> — not yet in the config reference)</li></ul> https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-0-43-1 2026-09-15T09:00:00Z <ul><li><b>CHG</b> <code>/copy</code> — native clipboard write path for Linux X11 sessions</li><li><b>FIX</b> <code>Ctrl-C</code> while subagents are running now interrupts only the subagents, instead of exiting the whole CLI</li></ul> https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-0-43-0 2026-09-14T09:00:00Z <ul><li><b>ADD</b> <code>/sessions</code> — <code>Ctrl-X</code> then <code>y</code> deletes the selected session from the picker</li><li><b>ADD</b> <code>kimi upgrade</code> / <code>kimi update</code> — <code>-y, --yes</code> skips the confirmation prompt and installs directly</li><li><b>ADD</b> <code>loop_control.compaction_max_attempts</code> — caps total attempts for a failing compaction request; default 5</li><li><b>ADD</b> <code>KIMI_CODE_PERMISSION_MODE_REMINDER</code> env var — set to <code>0</code> to stop injecting auto permission-mode reminders into the model context</li><li><b>CHG</b> <code>Ctrl-S</code> — a steering message can now interrupt a wait for a background task</li><li><b>CHG</b> <code>/goal</code> — the time budget no longer counts time spent with the session closed, and the 24-hour limit is removed</li></ul> https://dommango.github.io/agentmats/kimi-code/changelog.html#rel-0-42-0 2026-09-09T09:00:00Z <ul><li><b>ADD</b> <code>card-keys</code> — initial reference: the seven keyboard-shortcut contexts, from general input to the approval panel (19 rows)</li><li><b>ADD</b> <code>card-slash-core</code> — initial reference: account, session, mode and status commands with their aliases (33 rows)</li><li><b>ADD</b> <code>card-slash-tools</code> — initial reference: swarm and goal control, extensibility, and the built-in skill commands (16 rows)</li><li><b>ADD</b> <code>card-cli</code> — initial reference: all 14 top-level flags and all 12 subcommands (25 rows)</li><li><b>ADD</b> <code>card-settings</code> — initial reference: <code>config.toml</code>, <code>tui.toml</code> and <code>.kimi-code/local.toml</code> (37 rows)</li><li><b>ADD</b> <code>card-env</code> — initial reference: core paths, model overrides, runtime switches and diagnostics (22 rows)</li><li><b>ADD</b> <code>card-skills</code> — initial reference: skill and agent directories, the four <code>AGENTS.md</code> locations, plugins (18 rows)</li><li><b>ADD</b> <code>card-hooks</code> — initial reference: 18 hook events, hook exit codes, <code>mcp.json</code> and MCP permissions (20 rows)</li><li><b>DEL</b> <code>KIMI_CODE_EXPERIMENTAL_REMOTE_CONTROL</code> and <code>KIMI_CODE_EXPERIMENTAL_SECONDARY_MODEL</code> — Remote Control and the subagent model pool are always on in 0.42.0, so both flags were removed</li><li><b>CHG</b> <code>[database]</code> replaces the experimental minidb flags; the read model and search worker are now controlled by <code>KIMI_CODE_PERSISTENCE_MINIDB_READMODEL</code> and <code>KIMI_CODE_SEARCH_WORKER</code></li><li><b>ADD</b> <code>[read]</code> — configurable character limits and resumable long-line file reads, replacing repeated output truncation</li><li><b>CHG</b> <code>Ctrl-O</code> — finished tool calls collapse to a header plus one outcome row; hidden output is counted and revealed here</li></ul>