The AI Toolchain — issue -152, March 20, 2026

AutoGPT Platform adds AgentMail blocks, AutoPilotBlock, GitHub CLI integration, graph search, and rich media previews in v0.6.52

$ git clone --branch autogpt-platform-beta-v0.6.52 https://github.com/Significant-Gravitas/AutoGPT.git
# already have the repo? check out this version:
$ git checkout autogpt-platform-beta-v0.6.52

• ›Adds AutoPilotBlock for invoking AutoPilot directly from graphs, enabling nested automation workflows.
• ›Adds GitHub CLI support by injecting GH_TOKEN and a connect_integration tool, enabling agent-driven GitHub operations.
• ›Adds AgentMail integration blocks, extending email automation surfaces available in the builder.
• ›Adds graph search functionality to the new builder, allowing practitioners to locate nodes across large agent graphs.
• ›Adds rich media previews for Builder node outputs and file inputs, surfacing artifacts inline during graph authoring.

Build resilient agents.

LangGraph CLI 0.4.19 adds deploy revisions list command to inspect deployment revisions.

$ git clone --branch cli==0.4.19 https://github.com/langchain-ai/langgraph.git
# already have the repo? check out this version:
$ git checkout cli==0.4.19

$ langgraph deploy revisions list

• ›Adds deploy revisions list subcommand to list revisions of a LangGraph deployment.

Semantic Kernel .NET 1.74.0 adds server URL validation for OpenAPI plugins, LINQ-based text search, and migrates DALL-E to gpt-image-1.

$ git clone --branch dotnet-1.74.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:
$ git checkout dotnet-1.74.0

• ›Adds AllowedDirectories deny-by-default configuration to DocumentPlugin to restrict file system access at the plugin level.
• ›Adds server URL validation options for OpenAPI plugins to control which remote endpoints plugins are permitted to call.
• ›Adds LINQ-based text search feature for .NET, enabling structured query expressions against text search providers.
• ›Migrates .NET image generation support from deprecated DALL-E models to gpt-image-1.
• ›Adds Python support for the new OpenAI text-to-image model.

$ git clone --branch vectordata-dotnet-10.1.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:
$ git checkout vectordata-dotnet-10.1.0

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 v2.1.81 adds --bare flag for headless scripting and --channels permission relay to mobile devices.

$ git clone --branch v2.1.81 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.81

$ claude --bare -p "Summarize the security findings in findings.json" --settings ci-settings.json

• ›Adds --bare flag for scripted -p calls — skips hooks, LSP, plugin sync, and skill directory walks; requires ANTHROPIC_API_KEY or apiKeyHelper via --settings (OAuth/keychain and auto-memory disabled).
• ›Adds --channels permission relay so channel servers that declare the permission capability can forward tool-approval prompts to your phone.
• ›Updates MCP OAuth to support Client ID Metadata Document (CIMD / SEP-991) for servers without Dynamic Client Registration.
• ›Improves MCP read/search tool calls to collapse into a single 'Queried {server}' line, expandable with Ctrl+O.
• ›Improves ! bash mode discoverability — Claude now suggests it when you need to run an interactive command.

Copilot CLI v1.0.10 adds /undo, multi-session support, --effort alias, SDK elicitation dialogs, and camelCase config keys.

$ git clone --branch v1.0.10 https://github.com/github/copilot-cli.git
# already have the repo? check out this version:
$ git checkout v1.0.10

$ /undo

{ "statusLine": true, "effortLevel": "high", "autoUpdatesChannel": "stable" }

• ›Adds --effort as a shorthand alias for --reasoning-effort flag.
• ›Adds /undo command to undo the last turn and revert file changes made during a session.
• ›Renames config settings to camelCase: includeCoAuthoredBy, effortLevel, autoUpdatesChannel, and statusLine (old snake_case names still accepted).
• ›Workspace MCP servers from .mcp.json, .vscode/mcp.json, and devcontainer.json are now loaded only after folder trust is confirmed.
• ›Plugins loaded via --plugin-dir now appear in /plugin list under a separate 'External Plugins' section.

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

Pi v0.61.1 adds typed tool_call handler return values and updated default models for zai, cerebras, and MiniMax providers.

$ git clone --branch v0.61.1 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.61.1

• ›Adds ToolCallEventResult type exports from @mariozechner/pi-coding-agent top-level and core extension entry points, enabling extension authors to strongly type tool_call handler return values.
• ›Updates default models for zai, cerebras, minimax, and minimax-cn providers and adds MiniMax-M2.1-highspeed model entries with normalized context limits.

$ git clone --branch v0.61.0 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.61.0

$ /export ~/sessions/pentest-2025-07-14.jsonl

$ /import ~/sessions/pentest-2025-07-14.jsonl

vLLM v0.18.0 adds gRPC serving, GPU-less render launch, NGram GPU speculative decoding, FlexKV offloading, and streaming tool calls in the Responses API.

$ git clone --branch v0.18.0 https://github.com/vllm-project/vllm.git
# already have the repo? check out this version:
$ git checkout v0.18.0

$ vllm launch render

$ vllm serve deepseek-ai/DeepSeek-V3 --enable-ep-weight-filter

• ›Adds --grpc flag to enable gRPC serving alongside the existing HTTP/REST interface for high-performance RPC-based inference.
• ›Adds vllm launch render subcommand for GPU-less multimodal preprocessing and rendering, separating preprocessing from GPU inference.
• ›Adds --enable-ep-weight-filter CLI option for faster Elastic Expert Parallelism model loading via NIXL-EP integration.
• ›Adds --distributed-timeout-seconds option for configuring distributed operation timeouts.
• ›Adds --attention-backend auto option for automatic attention backend selection.

vMLX - JANGTQ Uber Compressed MLX Models - L2 Disk Cache (survives restart) + L1 Paged (super fast ttft) + Hybrid SSM Scheduler + Cont Batching + etc!

vMLX v1.3.0 adds Nemotron-H support with hybrid SSM caching and session status banners with smooth token streaming.

$ git clone --branch v1.3.0 https://github.com/jjang-ai/vmlx.git
# already have the repo? check out this version:
$ git checkout v1.3.0

• ›Adds Nemotron-H JANG engine support with gate dequant (8→2 bit), fc1/fc2 rename, and MTP filter (42 GB GPU, 46 tok/s).
• ›Extends hybrid SSM full cache — prefix, paged, and disk caching — to Qwen3.5-A3B and Nemotron-H models.
• ›Adds session status banners showing loading, sleeping, and stopped states in chat UI.
• ›Adds renderer-side typewriter animation via requestAnimationFrame (rAF) for smooth token streaming.

Composio CLI 0.2.6 reorganizes commands into consumer vs. developer scopes and adds smarter execute defaults.

$ git clone --branch @composio/[email protected] https://github.com/ComposioHQ/composio.git
# already have the repo? check out this version:
$ git checkout @composio/[email protected]

• ›Moves composio search, composio link, and composio execute to consumer-only top-level commands, with developer-scoped operations now under composio manage ....
• ›Defaults composio execute to an empty object {} when no -d/--data flag or piped stdin is provided, eliminating the need to pass an explicit empty payload.
• ›Uses -d "{}" in the search call-to-action for tools with no schema properties, ensuring shell-safe invocation.
• ›Removes developer-only flags from root help output and adds short related-command hints pointing to composio manage ....