KimiThinkingPrefill

KO | EN

gitlite — search

#javascript #react #ai #typescript #android #linux #docker #macos #windows #vue #security #mcp #openai #open-source

KimiThinkingPrefill

(★ 10)

SillyTavern extension: client-side reasoning_content thinking prefill for Kimi/Moonshot models (kimi-k3-jb patch equivalent, no server modifications)

KimiThinkingPrefill 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Kimi Thinking Prefill

Client-side SillyTavern extension implementing the
[kimi-k3-jb patch](https://rentry.org/kimi-k3-jb) (`reasoning_content` thinking prefill for
Kimi/Moonshot models) without modifying any server files.

## What it does

Hooks `CHAT_COMPLETION_SETTINGS_READY` and rewrites the outgoing request payload:

1. **Injection** — when the prompt does not end on an assistant message, appends:

```json
   { "role": "assistant", "content": "", "reasoning_content": "<your prefill>", "partial": true }
   ```

The model then *continues thinking* from the prefilled reasoning, per the Moonshot API's
   documented partial-prefill behavior.

2. **Preset parity (patch transform)** — if the prompt already ends on an assistant message
   whose content starts with `<think>`, the think block is moved into `reasoning_content`
   and the message is flagged `partial: true` — the exact transform the server patch adds to
   `addAssistantPrefix`. This means preset-style prefills like
   `<think>I should continue the story.` keep working unchanged.

3. **Preserved thinking (optional)** — re-attaches each prior assistant message's stored reasoning
   (`chat[i].extra.reasoning`) as `reasoning_content` in the outgoing payload, so Kimi's
   [preserved-thinking behavior](https://platform.kimi.ai/docs/guide/use-thinking-models) works in
   multi-turn chats. Off by default; requires the SillyTavern **Show thoughts** toggle to be on
   (reasoning is only persisted then). Note: prior reasoning is billed as input tokens.

## Guards (mirroring the patch)

- Only runs when the current model id matches the configurable filter (default `kimi,moonshot`,
  covers `moonshotai/kimi-k3` on OpenRouter and `kimi-k3`/`kimi-k2-*` on the direct Moonshot API).
- Skipped when JSON schema / structured output is active.
- Skipped when tools/function calling are in play.
- Injection only on normal/regenerate/swipe generations (never quiet prompts or impersonation);
  the `<think>` transform additionally applies on Continue.

## Settings

Extensions menu → **Kimi Thinking Prefill**:

- **Enable thinking prefill** — toggle for the prefill features (transform + injection).
  Independent of the re-attach toggle below; either works without the other.
- **reasoning_content prefill** — the thinking text to prefill (plain text, no `<think>` tag needed).
- **Model filter** — comma-separated substrings matched against the model id.
- **Force thinking on for prefilled requests** (default: on) — sets `include_reasoning` on requests
  this extension modifies. **Required**: with thinking disabled the model continues the seeded
  `reasoning_content` with reply text and never reasons (reply shows up inside the reasoning panel).
- **Send all prior assistant reasoning back to the API** (default: off) — the preserved-thinking
  feature above. Pairs chat messages to outgoing assistant messages 1:1 (system messages excluded).
- **Log decisions to browser console** — debug output for each guarded decision.

## Verification

Enable debug logging, generate, and check the browser console for
`[KimiThinkingPrefill] Injected reasoning_content prefill: ...`. On the server side (ST terminal
with request logging), the final message should look like:

```json
{ "role": "assistant", "content": "", "reasoning_content": "I should continue the story.", "partial": true }
```

## Notes

- Works with the direct Moonshot source, OpenRouter (Moonshot provider), and Custom endpoints —
  anywhere the model id matches the filter and the API honors `partial`/`reasoning_content`.
- The server patch is *not* required; do not run both (double transforms are harmless but pointless).

KimiThinkingPrefill

KimiThinkingPrefill 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

KimiThinkingPrefill

KimiThinkingPrefill 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README