THE AI TOOLCHAIN NO. -216 The daily firehose — everything the toolchain shipped today, already filtered.
We read every release from the 174 tools on our watchlist at the source — GitHub and GitLab release notes, vendor release pages and changelogs, project blogs and feeds, vendor press releases, and the source code behind the tag. Bug-fix-only releases and non-product newsroom noise are dropped; what's left is summarized down to the new capability, how to try it, and any screenshots or videos the release itself published. Every entry links to the sources it was built from.
$ tct list # 9 tools matched
Agno v2.3.26 adds per-request isolation for agents, teams, and workflows in shared FastAPI processes.
$ git clone --branch v2.3.26 https://github.com/agno-agi/agno.git # already have the repo? check out this version: $ git checkout v2.3.26
Your own personal AI assistant. Any OS.
OpenClaw v2026.1.12 adds vector memory search, voice-call plugin parity, ISO-timestamp cron, new AI models, and Tailscale gateway support.
$ git clone --branch v2026.1.12 https://github.com/openclaw/openclaw.git # already have the repo? check out this version: $ git checkout v2026.1.12
$ clawdbot cron add --agent <agentId> --at 2026-02-01T09:00:00Z --delete-after-run "summarize overnight alerts"
# In your OpenClaw config:
tools:
sandbox:
tools:
- group: default
docker:
binds:
- /host/data:/workspace/data:ro clawdbot memory CLI with memory_search and memory_get tools, returning snippets and line ranges; index stored at ~/.clawdbot/memory/{agentId}.sqlite with watch enabled by default.node-llama-cpp embeddings, with per-agent overrides and provider metadata surfaced in tools and CLI.tools.sandbox.tools; allows Docker bind mounts via docker.binds.scrollintoview action in the tools layer; adds xhigh thinking support for GPT-5.2/Codex with safe downgrades.clawdbot dashboard auto-open flow.discord.allowBots config option; routes MiniMax vision to the Coding Plan VLM endpoint with @/path/to/file.png input support.ackMaxChars to 300 so short HEARTBEAT_OK replies stay internal.clawdbot doctor --non-interactive automatically after git installs and updates.channels.*.memory tool shorthand is dropped from tool profiles/group shorthands.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 v2.1.6 adds config search, date-ranged stats, nested skills discovery, and richer context window metrics.
$ git clone --branch v2.1.6 https://github.com/anthropics/claude-code.git # already have the repo? check out this version: $ git checkout v2.1.6
$ /stats # then press 'r' to cycle through Last 7 days / Last 30 days / All time $ # In your status line config, reference the new fields:
# context_window.used_percentage
# context_window.remaining_percentage /config command for quickly locating settings./doctor showing the auto-update channel and available npm versions (stable/latest)./stats — press r to cycle between Last 7 days, Last 30 days, and All time..claude/skills directories when working with files in subdirectories.context_window.used_percentage and context_window.remaining_percentage fields to the status line for easier context window display./tasks dialog to go directly to task details when only one background task is running.-mention MCP servers to enable/disable them has been removed — use /mcp enable <name> instead.GitHub Copilot CLI v0.0.381 adds --allow-all/--yolo permission flags, /new alias, and prefix-filtered shell history navigation.
$ git clone --branch v0.0.381 https://github.com/github/copilot-cli.git # already have the repo? check out this version: $ git checkout v0.0.381
$ gh copilot suggest --allow-all 'set up a new node project with eslint'
--allow-all and --yolo flags to enable all permissions at once without individual prompts./new as an alias for the /clear command in chat.!git and pressing up arrow cycles only through matching previous commands.GitHub Copilot CLI v0.0.380 adds interactive --agent mode, sub-agent abort propagation, large-file reads, and mid-thinking message queuing.
$ git clone --branch v0.0.380 https://github.com/github/copilot-cli.git # already have the repo? check out this version: $ git checkout v0.0.380
--agent flag now works in interactive mode, enabling agent-driven workflows without leaving the interactive shell.view_range parameter is specified in the view tool.web-fetch tool now rejects file:// URLs and redirects users to the view tool instead.The open source coding agent.
OpenCode v1.1.18 adds a fullscreen permission prompt view and provider icons in the desktop app.
$ git clone --branch v1.1.18 https://github.com/sst/opencode.git # already have the repo? check out this version: $ git checkout v1.1.18
OpenCode v1.1.16 adds CLI debug tool calling, desktop password-protected local server, and MCP server management fixes.
$ git clone --branch v1.1.16 https://github.com/sst/opencode.git # already have the repo? check out this version: $ git checkout v1.1.16
AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI
Pi v0.45.4 adds Vercel AI Gateway provider support and new extension examples for custom UI workflows.
$ git clone --branch v0.45.4 https://github.com/earendil-works/pi.git # already have the repo? check out this version: $ git checkout v0.45.4
$ AI_GATEWAY_API_KEY=<your-key> pi --provider vercel-ai-gateway
--provider vercel-ai-gateway and AI_GATEWAY_API_KEY environment variable.summarize.ts extension example for summarizing conversations using custom UI and an external model.questionnaire.ts extension example for multi-question input with tab bar navigation.plan-mode/ extension example with explicit step tracking and a progress widget.sharp to wasm-vips, eliminating native build requirements.Pi v0.45.0 adds MiniMax and Amazon Bedrock providers, model reordering in /scoped-models, and a bash sandbox extension example.
$ git clone --branch v0.45.0 https://github.com/earendil-works/pi.git # already have the repo? check out this version: $ git checkout v0.45.0
MINIMAX_API_KEY environment variable and minimax/MiniMax-M2.1 model identifier./scoped-models; order persists on Ctrl+S save and controls Ctrl+P cycling sequence.sandbox/ extension example demonstrating OS-level bash sandboxing via @anthropic-ai/sandbox-runtime with per-project configuration.Qwen Code SDK TypeScript v0.1.3 adds /skills command, Jupyter Notebook support, custom headers, shell timeouts, and sandbox customization.
$ git clone --branch sdk-typescript-v0.1.3 https://github.com/QwenLM/qwen-code.git # already have the repo? check out this version: $ git checkout sdk-typescript-v0.1.3
$ /skills /skills command with hot reload support for managing agent skills at runtime..ipynb) file code selection.Qwen Code v0.7.0-preview.0 adds multi-provider model config, an experimental /skills command with hot reload, Jupyter Notebook support, custom headers, and sandbox customization.
$ git clone --branch v0.7.0-preview.0 https://github.com/QwenLM/qwen-code.git # already have the repo? check out this version: $ git checkout v0.7.0-preview.0
$ /skills /skills command with hot reload, enabling in-session skill management without restarting the CLI..ipynb) file code selection, enabling context-aware interactions with notebook cells.Qwen Code v0.7.0 adds multi-provider model configuration support.
$ git clone --branch v0.7.0-nightly.20260113.09196c6e https://github.com/QwenLM/qwen-code.git # already have the repo? check out this version: $ git checkout v0.7.0-nightly.20260113.09196c6e
llama.cpp b7723 adds fattn-mma-f16 flash attention kernel for AMD RDNA4 GPUs via HIP.
$ git clone --branch b7723 https://github.com/ggml-org/llama.cpp.git # already have the repo? check out this version: $ git checkout b7723
fattn-mma-f16 flash attention kernel for RDNA4 GPUs via the HIP backend, enabling hardware-accelerated matrix multiply-accumulate attention on AMD RDNA4 hardware.llama-bench gains a direct_io parameter for benchmarking with direct I/O enabled.
$ git clone --branch b7718 https://github.com/ggml-org/llama.cpp.git # already have the repo? check out this version: $ git checkout b7718
direct_io parameter to llama-bench to control direct I/O during benchmarking runs.llama.cpp b7716 adds a server argument to disable prompt caching for deterministic OpenAI-compat responses.
$ git clone --branch b7716 https://github.com/ggml-org/llama.cpp.git # already have the repo? check out this version: $ git checkout b7716
Composio v0.10.6 adds Cloudflare Workers support, auto_upload_download_files flag, and Vercel AI v6 provider upgrade.
$ git clone --branch v0.10.6 https://github.com/ComposioHQ/composio.git # already have the repo? check out this version: $ git checkout v0.10.6
auto_upload_download_files boolean flag to the Composio() Python constructor to control automatic file upload/download behavior.composio.triggers.verifyWebhooks to support webhook versions v1, v2, and v3.