The daily firehose — everything the toolchain shipped today, already filtered.
// HOW THIS ISSUE IS MADE
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.
Agno v2.3.7 adds Amazon Redshift toolkit and revamps Human-in-the-Loop with a new RunRequirement class
└──▷ GET THIS VERSION
$ git clone --branch v2.3.7 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:$ git checkout v2.3.7
›Introduces the RunRequirement class to represent and resolve Human-in-the-Loop requirements; these objects appear in the Agent response or in the RunPaused event during streaming HITL flows.
›Adds yield_run_response parameter to continue_run streaming methods, yielding a RunOutput object at the end of a continued run.
›Adds Amazon Redshift toolkit for exploring Redshift databases and running queries.
›Passes run_context into get_relevant_documents_from_knowledge so custom knowledge retrievers now have access to dependencies.
›Enables Agno evals via AgentOS with Agents and Teams that use an asynchronous database class.
LangChain Core 1.1.1 adopts UUID v7 for run IDs, bringing time-ordered identifiers to traces and callbacks.
└──▷ GET THIS VERSION
$ git clone --branch langchain-core==1.1.1 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain-core==1.1.1
›Switches run ID generation to UUID v7, providing time-sortable identifiers for runs, traces, and callbacks.
LangChain 1.1.1 switches run IDs to UUID v7 for time-ordered, sortable trace identifiers.
└──▷ GET THIS VERSION
$ git clone --branch langchain==1.1.1 https://github.com/langchain-ai/langchain.git
# already have the repo? check out this version:$ git checkout langchain==1.1.1
›Run IDs now use UUID v7, enabling time-ordered, lexicographically sortable identifiers for traces and runs.
Continue CLI gains headless TTY-free operation and grep/find/findstr search fallback in v1.7.0
└──▷ GET THIS VERSION
$ git clone --branch @continuedev/[email protected] https://github.com/continuedev/continue.git
# already have the repo? check out this version:$ git checkout @continuedev/[email protected]
›Adds grep, find, and findstr as alternative search strategies in the CLI, giving the agent fallback file-search methods when the primary strategy is unavailable.
›Removes the TTY requirement for the CLI in headless mode, enabling unattended/CI pipeline use without a pseudo-terminal.
›Adds Supabase-specific OAuth scopes to the mcp-remote fallback path for MCP integrations.
›Switches Gemini and Vertex AI response streaming to the google/genai SDK.
Continue v1.35.0 auto-approves parallel read-only built-in tool calls without user prompts.
└──▷ GET THIS VERSION
$ git clone --branch @continuedev/[email protected] https://github.com/continuedev/continue.git
# already have the repo? check out this version:$ git checkout @continuedev/[email protected]
Pi v0.12.9 adds /copy command to grab the last agent message straight to your clipboard.
└──▷ GET THIS VERSION
$ git clone --branch v0.12.9 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:$ git checkout v0.12.9
└──▷ TRY IT
After receiving a long Markdown-formatted agent response, copy it to clipboard for pasting into a report or ticket without losing formatting.
$ /copy
›Adds /copy command to copy the last agent message to the clipboard, with cross-platform support (macOS, Windows, Linux) — useful for extracting text from rendered Markdown output.
Pi v0.12.7 adds context compaction with manual and auto modes, plus branch source tracking for session lineage.
└──▷ GET THIS VERSION
$ git clone --branch v0.12.7 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:$ git checkout v0.12.7
└──▷ TRY IT
Manually compact a long session mid-conversation, directing the summary to focus on a specific area of work.
$ /compact focus on the authentication module decisions and discard unrelated tangents
Enable auto-compaction so long-running agent sessions never stall due to context overflow.
$ /autocompact
›Adds /compact [instructions] command to manually summarize older messages and reduce context usage, with optional custom instructions for the summary.
›Adds /autocompact toggle to automatically compact context when it reaches contextWindow - reserveTokens (default 16k reserve tokens).
›Compacted sessions display a collapsible summary in the TUI (toggled with o key) and in HTML exports.
›RPC mode gains a {"type":"compact"} command and auto-compaction support, emitting compaction events.
›Branched sessions now record branchedFrom in the session header to track the originating session file path.
Qwen Code v0.4.0-preview.0 adds session-level conversation history management.
└──▷ GET THIS VERSION
$ git clone --branch v0.4.0-preview.0 https://github.com/QwenLM/qwen-code.git
# already have the repo? check out this version:$ git checkout v0.4.0-preview.0
›Adds session-level conversation history management, enabling practitioners to view and control conversation context within a session.
Qwen Code v0.4.0-nightly adds session-level conversation history management.
└──▷ GET THIS VERSION
$ git clone --branch v0.4.0-nightly.20251204.6729980b https://github.com/QwenLM/qwen-code.git
# already have the repo? check out this version:$ git checkout v0.4.0-nightly.20251204.6729980b
›Adds session-level conversation history management, enabling users to control and review conversation context within a session.
›Adds prefill support to chat completions via the /lcpp/v1/chat/completions endpoint (distinct from the OpenAI-compatible /v1/chat/completions), enabling 'Assistant Continue' partial-turn continuation for frontends that support it.
›Adds support for the Flux2 image generation model, loadable in GGUF format with a separate VAE (flux2vae.safetensors) and a Clip 1 text encoder.
›Adds support for Z-Image Turbo image generation, a lightweight alternative to Flux2, using z_image_turbo-Q4_0.gguf with the Flux1 VAE and a Qwen3-4B text encoder as Clip 1; a .kcppt template is available to auto-download all three components.
›Adds Mistral and Ministral 3 model support, merged from upstream llama.cpp.
›Adds a secondary terminal UI on Linux to view console logging, launchable from the Extras tab even when not started from CLI.
llama.cpp b7275 adds Metal GPU kernels for TRI, FILL, EXPM1, and SOFTPLUS ops on Apple Silicon.
└──▷ GET THIS VERSION
$ git clone --branch b7275 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b7275
›Adds Metal GPU kernel implementations for the TRI, FILL, EXPM1, and SOFTPLUS operations, enabling these ops to run natively on Apple Silicon GPUs.
└──▷ BREAKING ON UPGRADE
!Linux releases will switch from .zip archives to .tar.gz archives in an upcoming release — deployment scripts that unzip Linux builds will need to be updated.
Get up and running with Kimi-K2.6, GLM-5.2, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Ollama v0.13.2 enables flash attention by default for vision models and adds Qwen3-Next.
└──▷ GET THIS VERSION
$ git clone --branch v0.13.2 https://github.com/ollama/ollama.git
# already have the repo? check out this version:$ git checkout v0.13.2
└──▷ TRY IT
Run the new Qwen3-Next model for fast, parameter-efficient inference.
$ ollama run qwen3-next
›Adds Qwen3-Next model, the first in the Qwen3-Next series, optimized for parameter efficiency and inference speed.
›Enables flash attention by default for vision models (mistral-3, gemma3, qwen3-vl, and others), improving memory utilization and performance when processing image inputs.
Arize Phoenix Evals 2.7.0 adds support for prompt/template messages in evaluations.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-evals-v2.7.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-evals-v2.7.0
›Supports prompt/template messages in evals, enabling structured message-based inputs to evaluation pipelines.
LanceDB python-v0.26.0-beta.0 adds IVF SQ index support, HNSW aliases, and namespace server-side query.
└──▷ GET THIS VERSION
$ git clone --branch python-v0.26.0-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:$ git checkout python-v0.26.0-beta.0
›Adds IVF_SQ index support and HNSW aliases for vector index creation.
›Supports server-side query scoped to a namespace, enabling filtered search without pulling full table listings client-side.
└──▷ BREAKING ON UPGRADE
!macOS x86 (Intel) is no longer supported; wheels for that platform will not be published.
!Namespace operations now use namespace models directly — code calling namespace APIs with the old model types will need to be updated.
LanceDB v0.23.0-beta.0 adds IVF SQ index support, HNSW aliases, and server-side namespace queries.
└──▷ GET THIS VERSION
$ git clone --branch v0.23.0-beta.0 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:$ git checkout v0.23.0-beta.0
›Adds IVF_SQ index support and HNSW aliases for vector index creation.
›Supports server-side query execution scoped to a namespace.
└──▷ BREAKING ON UPGRADE
!Mac x86 (Intel) is no longer supported; macOS users must run on Apple Silicon or another supported platform.
!Namespace operations now use namespace models directly — code calling namespace APIs must be updated to use the new model-based interface.
THE AI TOOLCHAIN