The AI Toolchain — issue -178, February 20, 2026

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.

Claude Code v2.1.50 adds worktree isolation for agents, new hook events, claude agents CLI command, and 1M context controls.

$ git clone --branch v2.1.50 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.50

$ claude agents

isolation: worktree

$ CLAUDE_CODE_DISABLE_1M_CONTEXT=1 claude -p "summarize this repo"

• ›Adds WorktreeCreate and WorktreeRemove hook events for custom VCS setup and teardown during agent worktree isolation.
• ›Supports isolation: worktree in agent definitions, letting agents declaratively run in isolated git worktrees.
• ›New claude agents CLI command lists all configured agents.
• ›Adds startupTimeout configuration option for LSP servers.
• ›Adds CLAUDE_CODE_DISABLE_1M_CONTEXT environment variable to disable 1M context window support.

GitHub Copilot CLI v0.0.413 adds remote plugin sources, configurable status line, and alt-screen mode on by default under --experimental

$ git clone --branch v0.0.413 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v0.0.413

{
  "plugins": [
    {
      "name": "my-remote-plugin",
      "source": "https://github.com/my-org/my-copilot-plugin"
    }
  ]
}

• ›Enables alt-screen mode by default when running with the --experimental flag.
• ›Adds configurable status line support to display dynamic session information via custom shell scripts.
• ›Supports remote plugin sources (GitHub repos and git URLs) in marketplace.json plugin entries.
• ›Adds ctrl+insert keybinding to copy selected text in the alt-screen view.
• ›Renders session info in the main view footer in alt-screen mode.

LocalAI v3.12.0 adds Voxtral TTS backend, multi-modal realtime conversations, multi-GPU Diffusers, and a new dark/light UI theme.

$ git clone --branch v3.12.0 https://github.com/mudler/LocalAI.git
# already have the repo? check out this version:
$ git checkout v3.12.0

• ›Adds voxtral backend for high-quality text-to-speech, available via the model gallery.
• ›Adds experimental sd_embed-style prompt embedding support to the diffusers backend for richer image generation pipelines.
• ›Enables multi-GPU support for the diffusers backend, improving image generation throughput.
• ›Extends the realtime API to accept text, image, and audio conversation items in a single session.
• ›Adds a left navbar with dark/light theme variants to the web UI.

llama.cpp b8117 adds RISC-V Vector (RVV) vec dot kernels for iq1_s, iq1_m, iq2_s, iq3_s, tq1_0, and tq2_0 quantization types.

$ git clone --branch b8117 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8117

• ›Adds RVV (RISC-V Vector) vectorized dot-product kernels for quantization types iq2_s, iq3_s, tq1_0, tq2_0, iq1_s, and iq1_m, accelerating inference on RISC-V hardware with vector extensions.
• ›Adds VLEN-aware switching logic for RVV vec dot kernels, enabling correct dispatch across RISC-V implementations with different vector register lengths.

$ git clone --branch b8116 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8116

$ llama-quantize --dry-run model.gguf output.gguf Q4_K_M

$ git clone --branch b8110 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8110

$ git clone --branch b8107 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8107

$ git clone --branch b8106 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8106

browser-use 0.11.11 adds custom HTTP headers, a max-clickable-elements parameter, file-attachment structured output, and a new direct CLI mode.

$ git clone --branch 0.11.11 https://github.com/browser-use/browser-use.git
# already have the repo? check out this version:
$ git checkout 0.11.11

• ›Adds max_clickable parameter to the Agent to cap the number of clickable elements considered per step.
• ›Adds custom HTTP headers support to the browser agent, letting callers inject arbitrary request headers.
• ›Adds file attachments to structured output, enabling agents to return file references alongside typed results.
• ›Introduces a new direct CLI mode for driving browser-use tasks without a Python script wrapper.

Phoenix 13.3.0 adds a conciseness classification evaluator and removes the model inferences/embeddings UI.

$ git clone --branch arize-phoenix-v13.3.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v13.3.0

• ›Adds a conciseness classification evaluator to the evals library for scoring response brevity.
• ›Removes model inferences, dimensions, embeddings, and point-cloud UI surfaces from the Phoenix interface.

Langfuse v3.155.0 adds historical single-observation evals, batch dataset actions on events, and Gemini 2.5 Pro Preview pricing support.

$ git clone --branch v3.155.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.155.0

• ›Adds metricsVersion prop to dashboard components to enable switching to the v4 Beta metrics engine.
• ›Adds position and level as filterable dimensions in the trace events table.
• ›Adds historical/batched single-observation evaluation runs, enabling eval pipelines over past observation data.
• ›Adds a batch action to add events from the events table directly to a dataset.
• ›Adds pricing and model support for gemini-3.1-pro-preview in the model prices registry.

Qdrant v1.17.0 adds Relevance Feedback, audit logging, weighted RRF, cluster telemetry, and Qdrant Edge (in-process mode).

$ git clone --branch v1.17.0 https://github.com/qdrant/qdrant.git
# already have the repo? check out this version:
$ git checkout v1.17.0

$ curl -X PUT 'http://localhost:6333/collections/my_collection/points' \
  -H 'Content-Type: application/json' \
  -d '{
    "update_mode": "insert",
    "points": [
      {"id": 1, "vector": [0.1, 0.2, 0.3], "payload": {"label": "example"}}
    ]
  }'

• ›Adds update_mode parameter to upsert operations, accepting upsert, update, or insert to control insert/update behavior per request.
• ›Adds secondary API key configuration for zero-downtime key rotation in distributed clusters.
• ›Adds a dedicated HTTP port for the /metrics endpoint to support internal monitoring without exposing the main API port.
• ›Adds an API to list shard keys for collections using user-defined sharding.
• ›Adds Audit Access Logging for tracking access to the Qdrant API.

Weaviate v1.34.15 adds a debug endpoint for LSM bucket views, batch logic for text2vec-google, and recursive nested property resolution.

$ git clone --branch v1.34.15 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.34.15

• ›Adds a debug endpoint for holding consistent views on LSM buckets, aiding low-level storage diagnostics.
• ›Adds naive batch logic to the text2vec-google module, enabling batched embedding requests.
• ›Returns all nested object properties recursively when specified implicitly in a query.

$ git clone --branch v1.33.17 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.33.17