Commit Graph

14906 Commits

Author SHA1 Message Date
Sydney Runkle
bf7a4841b0 patch skipped 2025-12-12 10:51:58 -05:00
Mason Daugherty
13dd115d1d docs(anthropic): nit comments (#34314) 2025-12-12 10:33:23 -05:00
Mason Daugherty
75d365418b style(core): docs nit (#34312) 2025-12-12 10:33:14 -05:00
Mason Daugherty
2cff369cdc feat(anthropic): accept TypedDict for built-in tool types (#34279)
Widen `bind_tools` to accept `TypedDict` via `Mapping` so that users may
import and use Anthropic's built-in tool types:

```python
import subprocess

from anthropic.types.beta import BetaToolBash20250124Param
from langchain.tools import tool

tool_spec = BetaToolBash20250124Param(
    name="bash",
    type="bash_20250124",
    strict=True,
)

@tool(extras={"provider_tool_definition": tool_spec})
def bash(*, command: str, restart: bool = False, **kw):
    """Execute a bash command."""
    if restart:
        return "Bash session restarted"
    try:
        result = subprocess.run(
            command,
            shell=True,
            capture_output=True,
            text=True,
            timeout=30,
        )
        return result.stdout + result.stderr
    except Exception as e:
        return f"Error: {e}"

# Bind bash tool to your model
```

---------

Co-authored-by: Chester Curme <chester.curme@gmail.com>
2025-12-12 10:29:12 -05:00
Christophe Bornet
f5b6eecf72 refactor(standard-tests): improve VCR config (#33968)
Use of the fixture `_base_vcr_config` is deprecated with alternative
function `base_vcr_config()`
This way:
* we don't need to import `_base_vcr_config` seen as unused (which leads
to ruff violations PLC0414 and F811)
* we don't need to make a copy since a new dict is created at each
function invocation

Co-authored-by: Mason Daugherty <mason@langchain.dev>
2025-12-12 10:14:26 -05:00
Jacob Lee
a528ea1796 feat(openai): Use responses API if model is gpt-5.2-pro (#34306) 2025-12-12 10:11:15 -05:00
Paul
bf6a5eb122 fix(huggingface): Helper logic for init_chat_model with HuggingFace backend (#34259) 2025-12-12 10:05:16 -05:00
j3r0lin
5720dea41b fix(openai): handle missing 'text' key in responses API content blocks (#34198) 2025-12-12 09:39:12 -05:00
Mohammad Mohtashim
087107557f chore(ollama,groq): Filtering Parameters in bind_tools for Ollama and Groq (#34167) 2025-12-12 09:24:24 -05:00
dumko2001
05ba853548 fix(ollama): pop unsupported 'strict' argument in ChatOllama (#34114) 2025-12-12 09:13:11 -05:00
Christophe Bornet
3fb90666be chore(langchain): cleanup ruff config (#32810)
Co-authored-by: Mason Daugherty <mason@langchain.dev>
Co-authored-by: Sydney Runkle <54324534+sydney-runkle@users.noreply.github.com>
Co-authored-by: Sydney Runkle <sydneymarierunkle@gmail.com>
2025-12-12 09:08:48 -05:00
Sydney Runkle
6a2a149f89 fix: little lint thing (#34310)
to be merged into https://github.com/langchain-ai/langchain/pull/32810
2025-12-12 08:47:51 -05:00
Christophe Bornet
bbc1d46efe chore(langchain): check agents integration tests with mypy (#34308) 2025-12-12 07:55:34 -05:00
Mason Daugherty
d6b5f05f33 refactor(anthropic): comments and _BUILTIN_TOOL_PREFIXES (#34305) 2025-12-11 16:57:22 -05:00
Mason Daugherty
10377a7373 fix(core): widen openai tool/function conversion input type to Mapping (#34304)
Motivated by changes to accept `TypedDict` tool types (e.g. in case of
Anthropic/Claude built-in tools)
2025-12-11 16:33:53 -05:00
ccurme
373ad8ac2c release(openai): 1.1.2 (#34302) langchain-openai==1.1.2 2025-12-11 16:20:57 -05:00
Mason Daugherty
5eec11e2db docs(anthropic): fix line number highlighting (#34303) 2025-12-11 16:12:01 -05:00
Jacob Lee
badc0cf1b6 fix(openai): Allow temperature when reasoning is set to the string 'none' (#34298)
Co-authored-by: Chester Curme <chester.curme@gmail.com>
2025-12-11 15:57:04 -05:00
Mason Daugherty
3b7abdff96 feat(anthropic): auto-apply mcp beta header (#34301)
and update docstring example
2025-12-11 15:49:32 -05:00
Mason Daugherty
4aebfbad59 docs(anthropic): use named betas param in docstring example (#34300) 2025-12-11 15:48:13 -05:00
Mason Daugherty
ae1f03fbe0 docs(anthropic): cleanup nits (#34299) 2025-12-11 15:17:56 -05:00
ccurme
46dbb3967e chore(anthropic): update test_tool_search cassette (#34297) 2025-12-11 10:53:52 -05:00
Mason Daugherty
dd0b990ba5 chore(infra): delete copilot instructions (#34294)
and some files we inherit from org root
2025-12-11 01:51:00 -05:00
ccurme
5aa46501cf fix(langchain): add sentinel value to ProviderStrategy / strict (#34290) 2025-12-10 16:25:06 -05:00
ccurme
92df109dd5 chore(langchain): add end to end test for strict mode in provider strategy (#34289) 2025-12-10 15:48:47 -05:00
Towseef Altaf
d27fb0c432 feat(langchain,openai): add strict flag to ProviderStrategy structured output (#34149) 2025-12-10 15:35:23 -05:00
ccurme
69dd39c461 fix(anthropic): ignore null values of caller on tool_use blocks (#34286) 2025-12-10 13:13:02 -05:00
ccurme
41cebfe4fb chore(core): add admonitions around use of load (#34285) 2025-12-10 11:36:46 -05:00
ccurme
5350967ddc feat(anthropic): support mcp_toolset in bind_tools (#34284) 2025-12-10 14:39:35 +00:00
Mason Daugherty
7542278997 feat(core,anthropic): extras on BaseTool (#34120) 2025-12-10 09:37:14 -05:00
Mason Daugherty
ff6e3558d7 docs(fireworks,groq,huggingface,mistralai,ollama,openai): x-ref convert_to_openai_tool (#34276) 2025-12-09 19:51:04 -05:00
Mason Daugherty
585e12e53b chore(infra): delete SECURITY.md (#34270)
Will be inherited from `langchain-ai/.github`
2025-12-09 15:01:53 -05:00
Sydney Runkle
73ba156a7d release: langchain-core 1.1.3 (#34266) langchain-core==1.1.3 2025-12-09 14:50:53 +00:00
Eugene Yurtsev
395c8d0bd4 fix(core): undo jinja2 restrictions (#34072)
Reverting jinja2 restrictions that made the feature unusable
2025-12-09 09:46:36 -05:00
Sydney Runkle
34d31b8394 fix: remove partial usage for retriever func + afunc (#34265)
Added test that fails on `master`.

`ToolNode` uses `get_type_hints` which doesn't work properly w/ partial
funcs on Python 3.12+

The diff here is nice anyways when we inline the logic.
2025-12-09 14:43:14 +00:00
Eugene Yurtsev
2aa0555941 chore(infra): update security.md file (#34258)
Move to github security features for intake

---------

Co-authored-by: Lauren Hirata Singh <lauren@langchain.dev>
2025-12-08 21:47:55 +00:00
Mason Daugherty
dff229d018 fix(openai): add missing tools param to ChatOpenAI with_structured_output (#34075) 2025-12-08 15:47:31 -05:00
Mason Daugherty
b009ca4d23 feat(standard-tests): invocation model override (#34170)
inspired by noticing `ChatGoogleGenerativeAI` failed to do so
2025-12-08 15:44:22 -05:00
Mason Daugherty
0254c12cb0 feat(standard-tests): ensure only one chunk sets model_name in usage_metadata (#34224) 2025-12-08 15:41:39 -05:00
Mason Daugherty
2faed37ff1 feat(anthropic): document and test fine grained tool streaming (#34118)
https://platform.claude.com/docs/en/agents-and-tools/tool-use/fine-grained-tool-streaming
2025-12-08 15:34:56 -05:00
Mason Daugherty
d886dcfba5 fix(standard-tests)!: remove deprecated has_tool_choice property (#34174)
Deprecated since `0.3.15`

This was marked as being removed in `0.3.20` but never was
2025-12-08 15:31:55 -05:00
Mason Daugherty
91d5ca275d feat(anthropic): use model profile for max output tokens (#34163)
Need(?) to adjust tests to also pull from model profile? currently
hardcoded
2025-12-08 15:31:16 -05:00
Mason Daugherty
dcb670f395 feat(anthropic): auto append relevant beta headers for computer use (#34117)
in addition to documenting it


https://platform.claude.com/docs/en/agents-and-tools/tool-use/computer-use-tool
2025-12-08 15:25:36 -05:00
ccurme
85012ae601 chore(infra): update default lib on release workflow (#34256) 2025-12-08 14:35:43 -05:00
ccurme
aa0f4fb927 release(langchain): 1.1.3 (#34255) langchain==1.1.3 2025-12-08 14:29:40 -05:00
Sydney Runkle
d18cdc6f32 feat: add agent name to AIMessage (#34254) 2025-12-08 14:23:12 -05:00
Mason Daugherty
8a5f46322b feat(anthropic): tool search support (#34119) 2025-12-08 10:46:37 -05:00
Mason Daugherty
a0e86b18bf release(core): 1.1.2 (#34253)
and bump deps
langchain-core==1.1.2
2025-12-08 10:24:03 -05:00
Nhan Nguyen
6affec92ce fix(core): pass tool_call_id to on_tool_start callback (#34235)
## Summary

When invoking a tool with a `ToolCall`, the `tool_call_id` is extracted
but was **not forwarded** to callback handlers in `on_tool_start`. This
made it impossible for callback handlers to correlate tool executions
with the original LLM tool calls.

This fix adds `tool_call_id=tool_call_id` to both:
- Sync `run()` method's `on_tool_start` call
- Async `arun()` method's `on_tool_start` call

## Changes

- **`libs/core/langchain_core/tools/base.py`**: Added `tool_call_id`
parameter to `on_tool_start` calls (2 lines)
- **`libs/core/tests/unit_tests/test_tools.py`**: Added 6 comprehensive
tests covering:
  - Sync tool invocation via `invoke()`
  - Async tool invocation via `ainvoke()`
  - `tool_call_id` is `None` when invoked without a ToolCall
  - Empty string `tool_call_id` edge case
  - Direct `run()` method
  - Direct `arun()` method

## Test plan

- [x] All 147 existing tests pass
- [x] 6 new tests added and passing
- [x] Linting passes

Fixes #34168

---

This PR was developed with AI assistance (Claude).

---------

Co-authored-by: Mason Daugherty <github@mdrxy.com>
Co-authored-by: Mason Daugherty <mason@langchain.dev>
2025-12-08 10:15:18 -05:00
Christophe Bornet
a64aee310c chore(core): improve typing of messages utils functions (#34225)
With this we get the correct types for `_runnable_support` annotated
functions.
* return list[BaseMessage] when messages is not None
* return Runnable when messages is None
* typing of function args
2025-12-08 09:59:43 -05:00