The AI Toolchain — issue -252, December 8, 2025

Glamourous agentic coding for all

Crush v0.22.0 adds DuckDuckGo web search and image-returning tool/MCP support via agentic fetch.

$ git clone --branch v0.22.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.22.0

• ›Adds DuckDuckGo-powered web search: tell Crush what you're looking for and it uses agentic fetch to search and retrieve results while keeping search context out of the main session.
• ›Supports image results returned by tools and MCPs (e.g., Playwright), enabling visual web agentic workflows.

Haystack v2.21.0 adds Multi-Query RAG components and async support for FilterRetriever and AutoMergingRetriever.

$ git clone --branch v2.21.0 https://github.com/deepset-ai/haystack.git
# already have the repo? check out this version:
$ git checkout v2.21.0

from haystack.components.query import QueryExpander
from haystack.components.retrievers import InMemoryBM25Retriever, MultiQueryTextRetriever
from haystack.document_stores.in_memory import InMemoryDocumentStore
from haystack.components.writers import DocumentWriter
from haystack import Document
from haystack.document_stores.types import DuplicatePolicy

store = InMemoryDocumentStore()
writer = DocumentWriter(document_store=store, policy=DuplicatePolicy.SKIP)
writer.run(documents=[Document(content="Renewable energy comes from wind and sunlight.")])

expander = QueryExpander()
retriever = InMemoryBM25Retriever(document_store=store, top_k=3)
multi_retriever = MultiQueryTextRetriever(retriever=retriever)

expanded = expander.run(query="renewable energy")
results = multi_retriever.run(queries=expanded["queries"])
for doc in results["documents"]:
    print(doc.content)

• ›Adds QueryExpander component (importable from haystack.components.query) to generate semantically similar query variations for broader search coverage.
• ›Adds MultiQueryTextRetriever (importable from haystack.components.retrievers) to run multiple queries in parallel against a text-based retriever (e.g., BM25) and merge results by score.
• ›Adds MultiQueryEmbeddingRetriever (importable from haystack.components.retrievers) to perform multi-query retrieval using embeddings for richer semantic recall.
• ›Adds return_empty_on_no_match parameter to RegexTextExtractor.__init__() (default True); set to False to return {'captured_text': ''} instead of {} when no regex match is found, ensuring consistent output structure for pipeline integration.
• ›FilterRetriever and AutoMergingRetriever components now support asynchronous execution.

LangChain 1.1.3 adds agent name to AIMessage, Anthropic effort support, and Upstage Solar in init_chat_model.

$ git clone --branch langchain==1.1.3 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain==1.1.3

• ›Adds effort parameter support for Anthropic models.
• ›Adds Upstage (Solar) as a supported provider in init_chat_model.
• ›Adds agent name to AIMessage objects.

$ git clone --branch langchain-core==1.1.2 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-core==1.1.2

llama.cpp b7318 lets you configure cache_reuse per request on the server.

$ git clone --branch b7318 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b7318

• ›Adds cache_reuse as a per-request configurable field on the llama.cpp server, allowing fine-grained control over KV-cache reuse on a request-by-request basis.

Langfuse v3.138.0 adds a DELETE API for prompts, SSO provider visibility in org tables, and improved OpenTelemetry/Pydantic tracing support.

$ git clone --branch v3.138.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.138.0

$ curl -X DELETE https://<your-langfuse-host>/api/public/prompts/<prompt-name> \
  -H 'Authorization: Basic <base64-encoded-credentials>'

• ›Adds DELETE endpoint for prompts, enabling programmatic deletion of prompt versions via the API.
• ›Adds SSO provider column to the organization members table in the UI, surfacing which SSO provider each member uses.
• ›Maps OpenTelemetry observation types for gen_ai (e.g. Pydantic AI) spans, improving OTEL ingestion compatibility.
• ›Renders Pydantic tool calls in a structured, human-readable format in the trace view.
• ›Adds trace_name column to the event table definition, expanding queryable fields on events.