@dreb/coding-agent 2.43.0 → 2.43.3

package/README.md

@@ -129,6 +129,8 @@ See [docs/providers.md](docs/providers.md) for detailed setup instructions, incl
 
 **Custom providers & models:** Add providers via `~/.dreb/agent/models.json` if they speak a supported API (OpenAI, Anthropic, Google). For custom APIs or OAuth, use extensions. See [docs/models.md](docs/models.md) and [docs/custom-provider.md](docs/custom-provider.md).
 
+**Reasoning across model switches:** Exact-model signed, encrypted, or redacted reasoning state is replayed unchanged. Structured reasoning is portable only between models that share a provider and the `openai-completions` API when the destination accepts the source's recognized plain field (`reasoning_content`, `reasoning`, or `reasoning_text`). Other readable reasoning is retained as labelled plaintext in `<reformatted-pre-switch-reasoning>` markers with incompatible protocol metadata removed; opaque redacted or encrypted-only state is omitted for incompatible targets. This outbound conversion does not change session history, so switching back can replay the original state unless it was compacted or pruned. Custom models must also share provider identity and compatible API/signature behavior.
+
 ---
 
 ## Interactive Mode

package/docs/custom-provider.md

@@ -608,7 +608,7 @@ interface ProviderModelConfig {
     requiresToolResultName?: boolean;
     requiresAssistantAfterToolResult?: boolean;
     requiresThinkingAsText?: boolean;
-    thinkingFormat?: "openai" | "openrouter" | "zai" | "qwen" | "qwen-chat-template";
+    thinkingFormat?: "openai" | "openrouter" | "zai" | "qwen" | "qwen-chat-template" | "kimi";
     openRouterRouting?: OpenRouterRouting;
     vercelGatewayRouting?: VercelGatewayRouting;
     supportsStrictMode?: boolean;

package/docs/models.md

@@ -11,6 +11,7 @@ Add custom providers and models (Ollama, vLLM, LM Studio, proxies) via `~/.dreb/
 - [Model Configuration](#model-configuration)
 - [Overriding Built-in Providers](#overriding-built-in-providers)
 - [Per-model Overrides](#per-model-overrides)
+- [Reasoning Across Model Switches](#reasoning-across-model-switches)
 - [OpenAI Compatibility](#openai-compatibility)
 
 ## Minimal Example
@@ -236,6 +237,12 @@ Behavior notes:
 - You can combine provider-level `baseUrl`/`headers` with `modelOverrides`.
 - If `models` is also defined for a provider, custom models are merged after built-in overrides. A custom model with the same `id` replaces the overridden built-in model entry.
 
+## Reasoning Across Model Switches
+
+A custom model's `provider` identity is part of reasoning-state compatibility; matching endpoint URLs or model IDs alone is not enough. Exact-model signed, encrypted, or redacted reasoning state is replayed unchanged. Between different models, structured reasoning is preserved only when both models use the same provider and `openai-completions` API, the destination accepts structured reasoning, and the source uses a recognized plain field: `reasoning_content`, `reasoning`, or `reasoning_text`.
+
+For other targets, readable reasoning is retained as labelled plaintext inside `<reformatted-pre-switch-reasoning>` markers after incompatible protocol metadata is stripped. Redacted or encrypted-only opaque state is omitted. This conversion happens only for the outbound request and does not alter session history, so returning to the original model can replay its original state unless history has been compacted or pruned. Portability also depends on compatible provider signatures.
+
 ## OpenAI Compatibility
 
 For providers with partial OpenAI compatibility, use the `compat` field.
@@ -270,7 +277,7 @@ For providers with partial OpenAI compatibility, use the `compat` field.
 | `requiresToolResultName` | Include `name` on tool result messages |
 | `requiresAssistantAfterToolResult` | Insert an assistant message before a user message after tool results |
 | `requiresThinkingAsText` | Convert thinking blocks to plain text |
-| `thinkingFormat` | Use `reasoning_effort`, `zai`, `qwen`, or `qwen-chat-template` thinking parameters |
+| `thinkingFormat` | Use `reasoning_effort`, `reasoning: { effort }`, `zai`, `qwen`, `qwen-chat-template`, or nested `thinking: { type, effort? }` parameters |
 | `supportsStrictMode` | Include the `strict` field in tool definitions |
 | `openRouterRouting` | OpenRouter routing config passed to OpenRouter for model/provider selection |
 | `vercelGatewayRouting` | Vercel AI Gateway routing config for provider selection (`only`, `order`) |

package/docs/providers.md

@@ -51,8 +51,10 @@ Use `/logout` to clear credentials. Tokens are stored in `~/.dreb/agent/auth.jso
 
 - `/login` uses the Kimi Code OAuth subscription endpoint at `https://api.kimi.com/coding/v1`.
 - `KIMI_API_KEY` uses Kimi For Coding's Anthropic-compatible API at `https://api.kimi.com/coding`.
-- The OAuth coding endpoint has been verified to accept OpenAI-style multimodal content arrays with base64 `image_url` data URLs for `kimi-for-coding`.
-- Moonshot Open Platform vision docs use a different base URL (`https://api.moonshot.ai/v1`), and the first-party Kimi CLI may have its own media handling; don't assume all Kimi routes expose identical media behavior.
+- Built-in OAuth models are `kimi-for-coding` (default, 262k context), `k3` (1M context), and `kimi-for-coding-highspeed` (262k context). Model availability is plan-dependent: the `k3` 1M-context ID and the `kimi-for-coding-highspeed` ID are only exposed when the subscription includes them, and `kimi-for-coding-highspeed` runs at roughly 6× speed for 3× quota usage. On login/refresh, dreb asks the Kimi API which models the subscription is entitled to and updates context, reasoning, image, tool-use, protocol, and thinking-effort metadata. Compatible newly discovered IDs are templated conservatively; if discovery fails, the static list remains available.
+- OAuth requests use the current Kimi Code device identity contract and share its stable `~/.kimi-code/device_id`. Extra low-precedence headers can be supplied with newline-separated `KIMI_CODE_CUSTOM_HEADERS` values.
+- The OAuth coding endpoint accepts OpenAI-style multimodal content arrays with base64 `image_url` data URLs. dreb keeps a conservative 32k output-token cap because the managed model catalog does not currently advertise a per-model output limit.
+- Moonshot Open Platform uses a different base URL (`https://api.moonshot.ai/v1`); don't assume both routes expose identical behavior.
 
 ## API Keys
 

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@dreb/coding-agent",
-	"version": "2.43.0",
+	"version": "2.43.3",
 	"description": "Coding agent CLI with read, bash, edit, write tools and session management",
 	"type": "module",
 	"drebConfig": {