i/next-ai-draw-io
1
0
Fork 0
mirror of https://github.com/DayuanJiang/next-ai-draw-io.git synced 2026-01-02 14:22:28 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat: Display AI reasoning/thinking blocks in chat interface (#152)

Browse Source 
* feat: Add reasoning/thinking blocks display in chat interface

* feat: add multi-provider options support and replace custom reasoning UI with AI Elements

* resolve conflicting reasoning configs and correct provider-specific reasoning parameters

* try to solve conflict

* fix: simplify reasoning display and remove unnecessary dependencies

- Remove Streamdown dependency (~5MB) - reasoning is plain text only
- Fix Bedrock providerOptions merging for Claude reasoning configs
- Remove unsupported DeepSeek reasoning configuration
- Clean up unused environment variables (REASONING_BUDGET_TOKENS, REASONING_EFFORT, DEEPSEEK_REASONING_*)
- Remove dead commented code from route.ts

Reasoning blocks contain plain thinking text and don't need markdown/diagram/code rendering.

* feat: comprehensive reasoning support improvements

Major improvements:
- Auto-enable reasoning display for all supported models
- Fix provider-specific reasoning configurations
- Remove unnecessary Streamdown dependency (~5MB)
- Clean up debug logging

Provider changes:
- OpenAI: Auto-enable reasoningSummary for o1/o3/gpt-5 models
- Google: Auto-enable includeThoughts for Gemini 2.5/3 models
- Bedrock: Restrict reasoningConfig to only Claude/Nova (fixes MiniMax error)
- Ollama: Add thinking support for qwen3-like models

Other improvements:
- Remove ENABLE_REASONING toggle (always enabled)
- Fix Bedrock providerOptions merging for Claude
- Simplify reasoning component (plain text rendering)
- Clean up unused environment variables

* fix: critical bugs and documentation gaps in reasoning support

Critical fixes:
- Fix Bedrock shallow merge bug (deep merge preserves anthropicBeta + reasoningConfig)
- Add parseInt validation with parseIntSafe helper (prevents NaN errors)
- Validate all numeric env vars with min/max ranges

Documentation improvements:
- Add BEDROCK_REASONING_BUDGET_TOKENS and BEDROCK_REASONING_EFFORT to env.example
- Add OLLAMA_ENABLE_THINKING to env.example
- Update JSDoc with accurate env var list and ranges

Code cleanup:
- Remove debug console.log statements from route.ts
- Refactor duplicate providerOptions assignments

---------

Co-authored-by: Dayuan Jiang <34411969+DayuanJiang@users.noreply.github.com>
Co-authored-by: Dayuan Jiang <jdy.toh@gmail.com>
This commit is contained in:
Biki Kalita
2025-12-10 20:54:43 +05:30
committed by GitHub
parent d2ba133eaf
commit a047a6ff97
10 changed files with 959 additions and 61 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
env.example
View File

@@ -11,28 +11,45 @@ AI_MODEL=global.anthropic.claude-sonnet-4-5-20250929-v1:0
# AWS_REGION=us-east-1
# AWS_ACCESS_KEY_ID=your-access-key-id
# AWS_SECRET_ACCESS_KEY=your-secret-access-key
# Note: Claude and Nova models support reasoning/extended thinking
# BEDROCK_REASONING_BUDGET_TOKENS=12000 # Optional: Claude reasoning budget in tokens (1024-64000)
# BEDROCK_REASONING_EFFORT=medium # Optional: Nova reasoning effort (low/medium/high)
# OpenAI Configuration
# OPENAI_API_KEY=sk-...
# OPENAI_BASE_URL=https://api.openai.com/v1 # Optional: Custom OpenAI-compatible endpoint
# OPENAI_ORGANIZATION=org-... # Optional
# OPENAI_PROJECT=proj_... # Optional
# Note: o1/o3/gpt-5 models automatically enable reasoning summary (default: detailed)
# OPENAI_REASONING_EFFORT=low # Optional: Reasoning effort (minimal/low/medium/high) - for o1/o3/gpt-5
# OPENAI_REASONING_SUMMARY=detailed # Optional: Override reasoning summary (none/brief/detailed)
# Anthropic (Direct) Configuration
# ANTHROPIC_API_KEY=sk-ant-...
# ANTHROPIC_BASE_URL=https://your-custom-anthropic/v1
# ANTHROPIC_THINKING_TYPE=enabled # Optional: Anthropic extended thinking (enabled)
# ANTHROPIC_THINKING_BUDGET_TOKENS=12000 # Optional: Budget for extended thinking in tokens
# Google Generative AI Configuration
# GOOGLE_GENERATIVE_AI_API_KEY=...
# GOOGLE_BASE_URL=https://generativelanguage.googleapis.com/v1beta # Optional: Custom endpoint
# GOOGLE_CANDIDATE_COUNT=1 # Optional: Number of candidates to generate
# GOOGLE_TOP_K=40 # Optional: Top K sampling parameter
# GOOGLE_TOP_P=0.95 # Optional: Nucleus sampling parameter
# Note: Gemini 2.5/3 models automatically enable reasoning display (includeThoughts: true)
# GOOGLE_THINKING_BUDGET=8192 # Optional: Gemini 2.5 thinking budget in tokens (for more/less thinking)
# GOOGLE_THINKING_LEVEL=high # Optional: Gemini 3 thinking level (low/high)
# Azure OpenAI Configuration
# AZURE_RESOURCE_NAME=your-resource-name
# AZURE_API_KEY=...
# AZURE_BASE_URL=https://your-resource.openai.azure.com # Optional: Custom endpoint (overrides resourceName)
# AZURE_REASONING_EFFORT=low # Optional: Azure reasoning effort (low, medium, high)
# AZURE_REASONING_SUMMARY=detailed
# Ollama (Local) Configuration
# OLLAMA_BASE_URL=http://localhost:11434/api # Optional, defaults to localhost
# OLLAMA_ENABLE_THINKING=true # Optional: Enable thinking for models that support it (e.g., qwen3)
# OpenRouter Configuration
# OPENROUTER_API_KEY=sk-or-v1-...