The AI Toolchain — issue -353, August 28, 2025

Autonomous coding agent as an SDK, IDE extension, or CLI assistant.

Cline v3.26.6 adds Qwen Code CLI API support with OAuth and promotes GPT-5 to the recommended model list.

$ git clone --branch v3.26.6 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v3.26.6

• ›Adds Qwen Code CLI API provider support with OAuth authentication.
• ›Adds GPT-5 to the recommended models list.

Zed now bundles its own Gemini CLI binary and ACP agents respect always_allow_tool_actions.

$ git clone --branch v0.201.6 https://github.com/zed-industries/zed.git
# already have the repo? check out this version:
$ git checkout v0.201.6

{
  "agent_servers": {
    "gemini": {
      "ignore_system_version": false
    }
  }
}

• ›Adds agent_servers.gemini.ignore_system_version config key: set to false to make Zed search $PATH for Gemini CLI before falling back to its bundled private copy.
• ›ACP agents now respect the always_allow_tool_actions setting, allowing tool actions to be auto-approved without per-action prompts.
• ›Zed now downloads and uses a private copy of the Gemini CLI binary by default, removing the dependency on a system-installed Gemini CLI.

An open-source AI agent that brings the power of Gemini directly into your terminal.

gemini-cli v0.3.0-preview.1 adds extension management commands, screen reader mode, prompt completion, fuzzy search toggle, and custom theme file paths.

$ git clone --branch v0.3.0-preview.1 https://github.com/google-gemini/gemini-cli.git
# already have the repo? check out this version:
$ git checkout v0.3.0-preview.1

$ gemini extensions install <extension-name>
gemini extensions list
gemini extensions uninstall <extension-name>

• ›Adds extensions install, uninstall, update, and list commands for managing CLI extensions.
• ›Introduces a screen reader mode flag for improved accessibility.
• ›Adds prompt completion support in the CLI REPL.
• ›Supports specifying themes as file paths instead of built-in names only.
• ›Adds an option to disable fuzzy search.

Framework for orchestrating role-playing, autonomous AI agents. By fostering collaborative intelligence, CrewAI empowers agents to work together seamlessly, tackling complex tasks.

CrewAI 0.175.0 adds Qdrant RAG support, crewai config reset, auto-injected trigger payloads, and a centralized embedding client system.

$ git clone --branch 0.175.0 https://github.com/crewAIInc/crewAI.git
# already have the repo? check out this version:
$ git checkout 0.175.0

$ crewai config reset

• ›Adds Qdrant as a RAG provider, with a new generic Qdrant client alongside an existing ChromaDB client.
• ›Introduces a centralized embedding types system and a base embedding client for standardized embedding configuration.
• ›Adds crewai config reset command to clear stored tokens.
• ›Enables crewai_trigger_payload auto-injection into flows triggered by automation.
• ›Adds support for additional parameters in Flow.start() methods.

langchain-text-splitters 0.3.10 adds optional custom header pattern support for text splitting.

$ git clone --branch langchain-text-splitters==0.3.10 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-text-splitters==0.3.10

• ›Adds optional custom header pattern support to the text splitter, allowing callers to supply their own regex or pattern definitions for header detection.

$ git clone --branch langchain==1.0.0a2 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain==1.0.0a2

Jan v0.6.9 adds mmproj/vision model support, MCP enhancements, GGUF validation, model favorites, and Polish localization.

$ git clone --branch v0.6.9 https://github.com/janhq/jan.git
# already have the repo? check out this version:
$ git checkout v0.6.9

• ›Adds support for passing the API key via environment variable instead of as a command-line argument for llama.cpp.
• ›Adds support for custom environment variables for llama.cpp, enabling fine-grained runtime configuration.
• ›Adds max_attempt setting for MCP to cap tool-call loops and prevent runaway execution.
• ›Adds a 'run on startup' setting for the local API server so it launches automatically with the app.
• ›Adds support for mmproj (multimodal projection) models, enabling vision/image capabilities in local inference.

Phoenix 11.30.0 adds a span document annotations API and missing-percentage visibility for experiment evaluations.

$ git clone --branch arize-phoenix-v11.30.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v11.30.0

• ›Adds a span document annotations API endpoint for annotating documents associated with spans.
• ›Experiments UI now shows the missing percentage for experiment evaluations, making gaps in eval coverage visible at a glance.

Langfuse v3.106.0 adds graph views for new observation types in tracing.

$ git clone --branch v3.106.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.106.0

• ›Adds graph views for new observation types in the tracing UI.

$ git clone --branch v3.105.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.105.0

Weave v0.52.5 adds user-controlled key redaction, a wandb integration, and parallel table uploads for faster tracing.

$ git clone --branch v0.52.5 https://github.com/wandb/weave.git
# already have the repo? check out this version:
$ git checkout v0.52.5

• ›Adds user control over which keys are redacted from traced payloads via feat(weave): Give user control over what keys to redact.
• ›Adds a wandb integration, enabling direct linking between Weave traces and W&B runs.
• ›Parallel table uploads via table_create_from_digests significantly reduce latency when logging large tables.
• ›Significantly improves import time by deferring imports until needed, reducing startup overhead for traced applications.
• ›Respects the WANDB_ENTITY environment variable in weave.init(), enabling entity selection without code changes.