The AI Toolchain — issue -050, June 30, 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 Sonnet 5 is now the default model in Claude Code, with a 1M-token context window and promotional pricing through August 31.

$ git clone --branch v2.1.197 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.197

• ›Switches default model to Claude Sonnet 5, featuring a native 1M-token context window for working across massive codebases in a single session.
• ›Introduces promotional pricing of $2/$10 per million input/output tokens through August 31, 2025.

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

Cline v4.0.5 adds Claude Sonnet 5 support across Anthropic, Bedrock, Vertex, and four more providers.

$ git clone --branch v4.0.5 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v4.0.5

• ›Supports Claude Sonnet 5 model across Anthropic, Bedrock, Vertex, Claude Code, SAP AI Core, OpenRouter, and Vercel AI Gateway providers.

GitHub Copilot CLI v1.0.67 adds Claude Sonnet 5 support, MCP OAuth fixes for Entra tenant vanity domains, and session resume hints.

$ git clone --branch v1.0.67 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.67

• ›Adds Claude Sonnet 5 as a supported model.
• ›MCP OAuth against Microsoft Entra servers behind a tenant vanity domain (e.g. Copilot Studio) now correctly refreshes and re-authenticates.
• ›Subagent sessions now inherit parent tool restrictions, preventing privilege-scope mismatches in nested agent calls.
• ›Disabling the sandbox mid-session now takes effect immediately, stopping shell and search commands from re-prompting to bypass it.
• ›Session limits are now required to be at least 30 AI credits.

$ git clone --branch v1.0.66 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.66

$ /worktree fix the login redirect

The open source coding agent.

OpenCode v1.17.12 adds yolo auto-approve mode, SDK event streams, session history, and desktop MCP autocomplete.

$ git clone --branch v1.17.12 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.17.12

• ›Adds a yolo mode to the TUI that auto-approves all permission prompts without interruption.
• ›New live event subscription stream in the SDK for real-time session monitoring.
• ›SDK now exposes access to active sessions programmatically.
• ›Adds paged durable session history via the SDK.
• ›New SDK endpoints for creating and fetching session permission requests.

AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

Pi v0.80.3 adds Claude Sonnet 5, Azure Foundry endpoints, RPC session inspection, and new UI/editor settings.

$ git clone --branch v0.80.3 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.80.3

{ "externalEditor": "code --wait" }

{ "outputPad": 2 }

• ›Adds Claude Sonnet 5 support via Anthropic-compatible and Bedrock provider catalogs with adaptive thinking enabled.
• ›Adds get_entries and get_tree RPC commands for inspecting session entries and tree snapshots over RPC.
• ›Adds a ./rpc-entry package export for launching Pi directly in RPC mode.
• ›Adds session_info_changed extension event so extensions can observe session name changes.
• ›Adds Azure OpenAI Responses provider support for modern Microsoft Foundry endpoint URLs.

LocalAI v4.5.6 adds native GGML voice/face-detect backends, semantic VAD EOU token for realtime, and distributed shared-model staging control via LOCALAI_DISTRIBUTED_SHARED_MODELS.

$ git clone --branch v4.5.6 https://github.com/mudler/LocalAI.git
# already have the repo? check out this version:
$ git checkout v4.5.6

• ›Adds LOCALAI_DISTRIBUTED_SHARED_MODELS environment variable to skip model staging on shared volumes in distributed deployments.
• ›Adds native voice-detect and face-detect GGML backends, replacing the previous Python insightface/speaker-recognition dependencies.
• ›Adds a Semantic VAD end-of-utterance (EOU) token to the realtime API.
• ›Introduces a SyncedMap component in distributed mode and migrates finetune, quantization, and agent-task state to cross-replica synchronisation.

llama.cpp b9844 adds NVFP4 support to the WebGPU backend.

$ git clone --branch b9844 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b9844

• ›Adds NVFP4 quantization support to the ggml-webgpu backend, enabling NVIDIA FP4 model inference via WebGPU.

Get up and running with Kimi-K2.6, GLM-5.2, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.

Gemma 4 on Apple Silicon now generates tokens ~90% faster via automatic multi-token prediction (MTP) — no config needed.

$ git clone --branch v0.31.1 https://github.com/ollama/ollama.git
# already have the repo? check out this version:
$ git checkout v0.31.1

• ›Enables multi-token prediction (MTP) for Gemma 4 on Apple Silicon, delivering ~90% faster token generation on average with no configuration or output changes.

Your own personal AI assistant. Any OS.

OpenClaw v2026.6.11 adds per-contact model assignment, Mattermost /oc_queue slash command, and QQBot slash-command scoping controls.

$ git clone --branch v2026.6.11 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:
$ git checkout v2026.6.11

$ /oc_queue

• ›Supports assigning different models to individual direct-message contacts across supported chat channels, while existing group and wildcard model choices continue to work.
• ›Adds /oc_queue native slash command for Mattermost operators to tune active-run queuing mode, debounce timing, cap, and drop handling directly in Mattermost.
• ›Enables QQBot group admins to control how broadly slash commands are available, with private-only commands directing users to a private chat instead of being exposed or silently ignored in groups.
• ›Mattermost thread participation now persists across gateway restarts without requiring users to re-mention the bot, surviving up to seven days of thread idle time.

Arize Phoenix client 2.11.0 adds an annotation summary view to project settings.

$ git clone --branch arize-phoenix-client-v2.11.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-client-v2.11.0

• ›Adds an annotation summary panel to project settings in the UI.

$ git clone --branch arize-phoenix-v17.14.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v17.14.0

$ git clone --branch arize-phoenix-v17.13.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v17.13.0

LanceDB python-v0.34.0-beta.5 adds OAuth connection config, Polars DataFrame integration, and monotonic reads via watermark header.

$ git clone --branch python-v0.34.0-beta.5 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout python-v0.34.0-beta.5

• ›Adds monotonic reads for remote tables via the x-lancedb-min-read-version watermark header, ensuring read-your-writes consistency in distributed scenarios.
• ›Exposes OAuth connection configuration for remote connections in both the Python and Node clients.
• ›Adds Polars DataFrame integration, enabling direct use of Polars DataFrames with LanceDB tables.
• ›Adds improved branch-handling capabilities to the LanceDB skill set for working with table branches.

$ git clone --branch v0.31.0-beta.5 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout v0.31.0-beta.5