The OpenAI Codex provider accepts reasoning_effort but silently
discards it. Wire it through as {"reasoning": {"effort": ...}} in
the request body so the config option actually takes effect.
- Remove trailing whitespace and normalize blank lines
- Unify string quotes and line breaks for long lines
- Sort imports alphabetically across modules
The bare keyword "codex" causes false positive matches when any model
name happens to contain "codex" (e.g. "gpt-5.3-codex" on a custom
provider). This incorrectly routes the request through the OAuth-based
OpenAI Codex provider, producing "OAuth credentials not found" errors
even when a valid custom api_key and api_base are configured.
Keep only the explicit "openai-codex" keyword so that auto-detection
requires the canonical prefix. Users can still set provider: "custom"
to force the custom endpoint, but auto-detection should not collide.
Closes#1311
_sanitize_messages strips all non-standard keys from messages, including
reasoning_content. Thinking-enabled models like Moonshot Kimi k2.5
require reasoning_content to be present in assistant tool call messages
when thinking mode is on, causing a BadRequestError (#1014).
Add reasoning_content to _ALLOWED_MSG_KEYS so it passes through
sanitization when present.
Fixes#1014
PR #947 fixed the consumer side (context.py) but the root cause is at
the provider level — getattr returns "" (empty string) instead of None
when reasoning_content is empty. This causes DeepSeek API to reject the
request with "Missing reasoning_content field" error.
`"" or None` evaluates to None, preventing empty strings from
propagating downstream.
Fixes#946
When MCP tools return empty content, messages may contain empty-string
text blocks. OpenAI-compatible providers reject these with HTTP 400.
Changes:
- Add _prevent_empty_text_blocks() to filter empty text items from
content lists and handle empty string content
- For assistant messages with tool_calls, set content to None (valid)
- For other messages, replace with '(empty)' placeholder
- Only copy message dict when modification is needed (zero-copy path
for normal messages)
Co-Authored-By: nanobot <noreply@anthropic.com>
Inject cache_control: {"type": "ephemeral"} on the system message and
last tool definition for providers that support prompt caching. Adds
supports_prompt_caching flag to ProviderSpec (enabled for Anthropic only)
and skips caching when routing through a gateway.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add VolcEngine ProviderSpec entry in registry.py
- Add volcengine to ProvidersConfig class in schema.py
- Update model providers table in README.md
- Add description about VolcEngine coding plan endpoint
- Strip non-standard keys like 'reasoning_content' before sending to LLM
- Always include 'content' key in assistant messages (required by StepFun)
- Add _sanitize_messages to LiteLLMProvider to prevent 400 BadRequest errors
Add SiliconFlow (硅基流动) as an OpenAI-compatible gateway provider.
SiliconFlow hosts multiple models (Qwen, DeepSeek, etc.) via an
OpenAI-compatible API at https://api.siliconflow.cn/v1.
Changes:
- Add ProviderSpec for siliconflow in providers/registry.py
- Add siliconflow field to ProvidersConfig in config/schema.py
Co-authored-by: Cursor <cursoragent@cursor.com>
