The AI Toolchain — issue -125, April 16, 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 v2.1.111 adds /ultrareview for cloud-based PR analysis, xhigh effort tier, PowerShell support, and a permission-prompt reducer.

$ git clone --branch v2.1.111 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.111

$ /ultrareview 4821

$ /less-permission-prompts

$ OTEL_LOG_RAW_API_BODIES=1 claude --output-format stream-json '<prompt>'

• ›Adds xhigh effort level for Opus 4.7, sitting between high and max; selectable via /effort, --effort, or the model picker.
• ›New /ultrareview command runs comprehensive code review in the cloud using parallel multi-agent analysis — no arguments reviews the current branch; /ultrareview <PR#> fetches and reviews a specific GitHub PR.
• ›New /less-permission-prompts skill scans transcripts for common read-only Bash and MCP tool calls and proposes a prioritized allowlist for .claude/settings.json.
• ›Auto mode is now available for Max subscribers using Opus 4.7, and no longer requires the --enable-auto-mode flag.
• ›Adds PowerShell tool support on Windows (progressive rollout); opt in/out with CLAUDE_CODE_USE_POWERSHELL_TOOL; enable on Linux/macOS with CLAUDE_CODE_USE_POWERSHELL_TOOL=1 (requires pwsh on PATH).

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

Cline v3.79.0 adds Claude Opus 4.7, Azure Blob Storage, and globalSkills remote config support.

$ git clone --branch v3.79.0 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v3.79.0

• ›Adds Claude Opus 4.7 model support.
• ›Adds Azure Blob Storage as a storage provider.
• ›Adds inline value reuse in user-level remote-config discovery.

Glamourous agentic coding for all

Crush v0.59.0 adds built-in jq support with a dedicated skill, skills discovery in the sidebar, and Rio Terminal progress bar feedback.

$ git clone --branch v0.59.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.59.0

• ›Adds built-in jq as a bash builtin so agents can query JSON files without requiring jq installed on the host machine.
• ›Includes a built-in jq skill to instruct the model on how to invoke jq during agentic tasks.
• ›Shows available skills in the sidebar and landing page for improved discoverability.
• ›Enables progress bar feedback when running inside Rio Terminal.

GitHub Copilot CLI v1.0.30 adds /statusline command to customize the status bar and expands image paste to Ctrl+V and Meta+V.

$ git clone --branch v1.0.30 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.30

$ /statusline branch quota

• ›Adds /statusline command (with /footer alias) to customize which items appear in the status bar, supporting: directory, branch, effort, context window, and quota.
• ›Extends image paste from clipboard to both Ctrl+V and Meta+V across all platforms.

$ git clone --branch v1.0.29 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.29

servers:
  - name: my-mcp-server
    url: https://mcp.example.com/v1

$ git clone --branch v1.0.28 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.28

The open source coding agent.

OpenCode v1.4.7 adds xhigh adaptive reasoning for Claude Opus 4.7 and cross-workspace auth context propagation.

$ git clone --branch v1.4.7 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.4.7

$ opencode --agent <your-agent>

• ›Supports xhigh adaptive reasoning effort for Claude Opus 4.7, enabling deeper multi-step reasoning.
• ›Claude Opus 4.7 now surfaces summarized thinking output by default.
• ›Auth context is now shared across workspace sessions, so provider sign-in persists when switching workspaces.
• ›Sessions now automatically retry provider 5xx errors even when the provider SDK does not flag them as retryable.
• ›Desktop builds display a Beta or Dev badge in the title bar when running pre-release builds.

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

Pi v0.67.6 adds prompt template argument hints, a new extension hook for HTTP response inspection, and OSC 8 hyperlink rendering.

$ git clone --branch v0.67.6 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.67.6

---
name: summarize
argument-hint: <file> [max-words]
description: Summarize a file
---
Summarize the contents of {{file}} in at most {{max-words}} words.

• ›Adds argument-hint frontmatter field for prompt templates, displayed before the description in the / autocomplete dropdown using <angle> for required and [square] for optional arguments.
• ›New after_provider_response extension hook lets extensions inspect provider HTTP status codes and headers after each response is received and before stream consumption begins.
• ›Compact interactive startup header now shows loaded AGENTS.md files, prompt templates, skills, and extensions as a comma-separated list; press Ctrl+O to toggle the expanded view.
• ›Markdown links in assistant output render as OSC 8 hyperlinks on supporting terminals, with safe fallback to plain text on unknown terminals and tmux/screen.

$ git clone --branch v0.67.4 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.67.4

$ pi --no-context-files "Explain what this repo does"

import { loadProjectContextFiles } from 'pi';

const files = await loadProjectContextFiles(process.cwd());
console.log('Context files found:', files);

The agent that grows with you

Hermes v2026.4.16 adds Nous Tool Gateway: web search, image gen, TTS, and browser automation for Portal subscribers.

$ git clone --branch v2026.4.16 https://github.com/NousResearch/hermes-agent.git
# already have the repo? check out this version:
$ git checkout v2026.4.16

$ hermes model

$ hermes tools && hermes status

• ›New Nous Tool Gateway gives paid Nous Portal subscribers automatic access to web search (Firecrawl), image generation (FAL / FLUX 2 Pro), text-to-speech (OpenAI TTS), and browser automation (Browser Use) — no extra API keys required.
• ›Per-tool opt-in via use_gateway config key, with full visibility through hermes tools and hermes status.
• ›Runtime now prefers the gateway over direct API keys when both are present.

LangChain Core 1.3.0a3 adds invocation-param tracing, ContextOverflowError, multimodal token counting, XML buffer formatting, and more.

$ git clone --branch langchain-core==1.3.0a3 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:
$ git checkout langchain-core==1.3.0a3

from langchain_core.exceptions import ContextOverflowError

try:
    response = llm.invoke(very_long_messages)
except ContextOverflowError as e:
    print(f'Context limit exceeded: {e}')
    # truncate or summarize messages and retry

from langchain_core.messages import HumanMessage
from langchain_core.utils.token_counter import count_tokens_approximately

messages = [
    HumanMessage(content=[
        {"type": "text", "text": "What is in this image?"},
        {"type": "image_url", "image_url": {"url": "https://example.com/photo.jpg"}}
    ])
]

print(count_tokens_approximately(messages))

• ›Adds ContextOverflowError exception class, raised automatically by Anthropic and OpenAI integrations when context window is exceeded.
• ›Adds multimodal support to count_tokens_approximately, enabling approximate token counting for messages containing images and other non-text content.
• ›Adds tool schema token counting to count_tokens_approximately, so tool definitions are included in context estimates.
• ›Adds scaling by reported usage in count_tokens_approximately to calibrate approximate counts against actual model-reported token usage.
• ›Adds xml format option to get_buffer_string() for serializing conversation history as XML.

Your own personal AI assistant. Any OS.

OpenClaw v2026.4.15 adds Gemini TTS, GitHub Copilot embeddings, LanceDB cloud storage, and a Model Auth status card.

$ git clone --branch v2026.4.15 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:
$ git checkout v2026.4.15

plugins:
  entries:
    memory-core:
      config:
        dreaming:
          storage:
            mode: "inline"

• ›Adds Gemini text-to-speech support to the bundled google plugin, with voice selection, WAV reply output, and PCM telephony output.
• ›Adds a Model Auth status card in Control UI/Overview showing OAuth token health and provider rate-limit pressure, with expiry callouts, backed by a new models.authStatus gateway method.
• ›Adds GitHub Copilot as an embedding provider for memory search, with a dedicated Copilot embedding host helper for plugin reuse.
• ›Adds cloud storage support to memory-lancedb so durable memory indexes can run on remote object storage instead of local disk only.
• ›Adds experimental agents.defaults.experimental.localModelLean: true to drop heavyweight default tools (browser, cron, message) and reduce prompt size for weaker local-model setups.

PydanticAI v1.83.0 adds xAI tool support, FastMCP metadata injection, Bedrock/Anthropic prompt caching, and a graceful parallel-tool end strategy.

$ git clone --branch v1.83.0 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v1.83.0

agent = Agent(model=model, end_strategy='graceful')

agent = Agent(model=model, output_type=str | None)

• ›Adds XSearchTool and FileSearch support for the xAI provider.
• ›Adds metadata injection per tool call via FastMCPToolset.
• ›Adds prompt cache TTL support for the Bedrock provider.
• ›Adds automatic prompt caching support for Anthropic.
• ›Adds a 'graceful' end strategy for parallel tool calls.

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

Ollama v0.21.0 adds Hermes Agent, GitHub Copilot CLI integration, and Gemma 4 on Apple Silicon MLX.

$ git clone --branch v0.21.0 https://github.com/ollama/ollama.git
# already have the repo? check out this version:
$ git checkout v0.21.0

$ ollama launch hermes

• ›Adds ollama launch hermes to run the Hermes Agent, which automatically creates skills to better serve research and engineering workflows.
• ›Adds GitHub Copilot CLI as a supported integration in ollama launch, configurable alongside other coding agents in a single command.
• ›Supports running Gemma 4 via the MLX backend on Apple Silicon, including a text-only MLX runtime.
• ›Adds mixed-precision quantization to the MLX backend.
• ›Expands MLX backend with new op wrappers: Conv2d, Pad, activations, trig, masked SDPA, and RoPE-with-freqs.

llama.cpp b8815 adds Metal ROLL op implementation for Apple Silicon GPU acceleration.

$ git clone --branch b8815 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8815

• ›Implements the ROLL operation for the Metal backend, enabling GPU-accelerated tensor rolling on Apple Silicon and iOS devices.

$ git clone --branch b8814 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8814

$ git clone --branch b8813 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8813

Arize Phoenix v14.8.0 adds Azure managed identity authentication for PostgreSQL connections.

$ git clone --branch arize-phoenix-v14.8.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v14.8.0

• ›Adds Azure managed identity authentication support for PostgreSQL database connections.

$ git clone --branch arize-phoenix-v14.7.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v14.7.0

Chroma 1.5.8 adds sharding-aware compaction, configurable RPC timeouts, pod anti-affinity in Helm, and new WAL read levels.

$ git clone --branch 1.5.8 https://github.com/chroma-core/chroma.git
# already have the repo? check out this version:
$ git checkout 1.5.8

• ›Adds IndexAndBoundedWal and IndexAndAdaptiveWal read levels to clients for finer-grained WAL read control.
• ›Makes admin RPC timeout configurable via the config system.
• ›Makes compaction client gRPC timeout configurable.
• ›Adds pod anti-affinity support to StatefulSet Helm templates for improved workload distribution.
• ›Adds a fault injection control plane for testing resilience scenarios.

LanceDB python-v0.31.0-beta.6 adds nested namespace support for listing databases.

$ git clone --branch python-v0.31.0-beta.6 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout python-v0.31.0-beta.6

• ›Supports nested namespace operations when listing databases.

$ git clone --branch v0.28.0-beta.6 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout v0.28.0-beta.6

Weaviate v1.37.0 adds a native MCP server, BlobHash property type, collection export, extensible tokenizers, and drop-vector-index support.

$ git clone --branch v1.37.0 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.37.0

• ›Adds BlobHash property type that automatically stores blob data as hashes, greatly reducing disk space required for blob workloads.
• ›Introduces a native MCP (Model Context Protocol) server interface (preview), enabling AI agents such as Claude and IDEs to natively read and write to Weaviate without custom code, with hybrid search, RAG, and multi-tenancy support out of the box.
• ›Adds collection export to cloud backends (AWS, GCP, Azure) and filesystem with point-in-time exports, multi-node support, concurrent exports, multi-tenancy handling, cancellation, and observability; disabled by default and configured via environment variables including a default path env var.
• ›Adds a tokenizer endpoint and middleware integration for extensible tokenizers (Phase 1), bringing self-serve, multilingual tokenization with accent-insensitive processing options for text properties and custom stopword presets.
• ›Adds experimental support for dropping vector indices from existing collections via an alter-schema endpoint, allowing memory reclamation; endpoint can be disabled via an environment setting.