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.
$ git clone --branch v2026.5.12 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:$ git checkout v2026.5.12
└──▷ USE IT
Keep ACP sessions alive by falling back to a backup runtime when the primary backend is unavailable before any output is emitted.
yaml
# In your openclaw config
acp:
fallbacks:
- backend: secondary-runtime
- backend: tertiary-runtime
Pin a remote gateway's TLS certificate explicitly for node-mode sessions so fresh endpoints fail closed without out-of-band trust.
yaml
# In your openclaw config
gateway:
remote:
tlsFingerprint: "sha256:AABBCC..."
Set the auto-scroll preference in WebChat/Control UI to always follow streaming output.
📍In the Control UI or WebChat, open the scroll mode selector and choose "Always follow streaming output" to lock the view to new content as it arrives.
›Adds acp.fallbacks config so ACP turns can try configured backup runtime backends when the primary backend is unavailable before any output is emitted.
›Adds a persisted auto-scroll mode selector in Control UI/WebChat: always follow streaming output, stay near-bottom, or disable auto-scroll and use the New Messages button manually.
›Externalizes Amazon Bedrock, Bedrock Mantle, Slack, OpenShell sandbox, and Anthropic Vertex provider/plugin packages so core installs only pull AWS/Slack/Vertex SDK dependencies when those providers are explicitly installed.
›Requires approval before exposing pending Node pairing commands, capabilities, and permissions in Gateway, and refreshes the live approved surface when pairings change.
›Requires v4 Gateway clients and streams explicit chat deltaText/replace frames so SDK clients can consume assistant updates without local diffing.
+8 moreshow less
›Requires system TLS trust before pinning a first-use direct wss:// gateway certificate on macOS, and honors gateway.remote.tlsFingerprint as an explicit pin for remote node-mode sessions.
›Discovers provider plugins from setup.providers[].envVars credentials during provider discovery.
›Serializes and retries semantic config mutations centrally so concurrent commands can rebase safe changes instead of clobbering each other.
›Enforces inbound media size caps on download streams for Feishu, WhatsApp, and Line plugins.
›Limits install-time code safety scans to plugin-owned runtime entrypoints while keeping dependency manifest denylist checks.
›Classifies ACP spawn-child sessions as kind: "spawn-child" in openclaw sessions and status output.
›Reports agentRuntime.id: "acpx" with source: "session-key" for ACP control-plane session rows in openclaw sessions --json, openclaw status, and Gateway session RPC responses.
›Honors --no-git-update for existing git checkouts before resolving release refs, preventing pinned source installs from moving during reinstall.
└──▷ BREAKING ON UPGRADE
!Provider apiKey values are now resolved only through structured env SecretRefs (secrets.providers[id] / secrets.defaults); broad environment variable strings matching ^[A-Z_][A-Z0-9_]*$ are no longer inferred as provider credentials, which will break setups that relied on that implicit inference.
!Gateway now requires v4 clients and streams explicit deltaText/replace frames; older SDK clients that relied on local diffing of assistant updates will break.
$ git clone --branch python-1.42.0 https://github.com/microsoft/semantic-kernel.git
# already have the repo? check out this version:$ git checkout python-1.42.0
›Improves Model Context Protocol (MCP) support with unspecified enhancements to the MCP integration.
›Adds percent-encoding for OpenAPI path parameters in the OpenAPI connector.
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 v2.1.142 expands claude agents configuration flags and upgrades Fast mode to Opus 4.7 by default.
└──▷ GET THIS VERSION
$ git clone --branch v2.1.142 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:$ git checkout v2.1.142
└──▷ TRY IT
Pin fast mode to Opus 4.6 to keep consistent behavior across a team before adopting the new default.
$ CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE=1 claude --fast
›Adds --add-dir, --settings, --mcp-config, --plugin-dir, --permission-mode, --model, --effort, and --dangerously-skip-permissions flags to claude agents for fine-grained control over dispatched background sessions.
›Fast mode now defaults to Opus 4.7 (up from 4.6); set CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE=1 to pin back to 4.6.
›Plugins with a root-level SKILL.md and no skills/ subdirectory are now automatically surfaced as a skill.
›The /plugin details pane and claude plugin details command now display LSP servers provided by a plugin.
›/web-setup now warns before overwriting an existing GitHub App connection.
└──▷ BREAKING ON UPGRADE
!Fast mode now uses Opus 4.7 by default instead of Opus 4.6; workloads sensitive to model changes must set CLAUDE_CODE_OPUS_4_6_FAST_MODE_OVERRIDE=1 to preserve previous behavior.
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
Cline CLI v3.0.3 adds isolated git worktree task execution, bearer token auth for AWS Bedrock, and Ollama thinking control.
└──▷ GET THIS VERSION
$ git clone --branch cli-v3.0.3 https://github.com/cline/cline.git
# already have the repo? check out this version:$ git checkout cli-v3.0.3
└──▷ TRY IT
Resume an existing task in a fresh, isolated git worktree to safely try a different approach without touching your main working tree.
$ cline --taskId <taskId> --continue--worktree
Run a new task in an isolated worktree so experimental changes are sandboxed from your current branch.
$ cline --worktree "Refactor the authentication module"
›Adds --worktree flag to auto-create a fresh git worktree under ~/.cline/worktrees/ for isolated task execution, compatible with --taskId and --continue.
›Shows session status in the CLI history view and refreshes status rows in place while the standalone history TUI is open.
›Supports bearer token auth for AWS Bedrock and correctly maps AWS profiles when configuring the Bedrock gateway.
›Adds --thinking none support for Ollama models that ship with reasoning enabled by default.
Get up and running with Kimi-K2.6, GLM-5.2, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
Ollama v0.24 launches Codex App support with built-in browser, review mode, and reworked MLX sampler.
└──▷ GET THIS VERSION
$ git clone --branch v0.24.0 https://github.com/ollama/ollama.git
# already have the repo? check out this version:$ git checkout v0.24.0
└──▷ TRY IT
Launch the Codex App to work on coding tasks with parallel threads, worktree support, and git integration.
$ ollama launch codex-app
Roll back to a previous Codex App configuration after an upgrade.
$ ollama launch codex-app --restore
›Adds ollama launch codex-app to run OpenAI's Codex desktop experience with parallel thread support, built-in worktree management, and git functionality.
›Codex App includes a built-in browser for loading local servers and annotating pages directly to request code changes.
›Codex App includes a Review mode for leaving inline comments and iterating on code without leaving the workspace.
›Supports ollama launch codex-app --restore to revert to a previous Codex App configuration.
›Reworked MLX sampler improves generation quality on Apple Silicon.
llama.cpp b9158 adds RDNA3 tensor core support for flash attention and extends CDNA head size support to 256.
└──▷ GET THIS VERSION
$ git clone --branch b9158 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b9158
›Adds RDNA3 tensor core (MMA) support to the CUDA flash attention kernel, using FP16 accumulation with 32-unit tiles for head sizes evenly divisible by 32.
›Extends CDNA1 flash attention support to attention head sizes up to 256.
›Improves AMD matrix transposition efficiency for RDNA3 and RDNA4 via wider 32-unit tiles that align with the warp size of 32.
›Adds a new ggml_cuda_mma::data_layout entry to guard against accumulator data layout misuse introduced by the new RDNA3/4 tiling scheme.
›Tunes flash attention kernel parameters for RDNA3, RDNA4, and CDNA1 architectures.
SYCL backend gains Level Zero memory allocation to eliminate multi-GPU host RAM mirroring on Intel Arc systems.
└──▷ GET THIS VERSION
$ git clone --branch b9145 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b9145
└──▷ TRY IT
Disable Level Zero memory allocation at runtime to fall back to standard SYCL paths, useful for diagnosing allocation issues on Intel Arc multi-GPU systems.
›Adds GGML_SYCL_ENABLE_LEVEL_ZERO runtime environment variable (default 1) to switch the SYCL backend between Level Zero (zeMemAllocDevice) and standard SYCL (sycl::malloc_device) memory APIs, eliminating the 1:1 VRAM-to-host-RAM mirroring that caused OOM crashes on multi-GPU Intel Arc systems.
›Adds GGML_SYCL_SUPPORT_LEVEL_ZERO CMake build option (default ON) to compile in Level Zero support when the SDK headers and loader are present, with automatic fallback to the SYCL allocation path when Level Zero is unavailable.
›Adds Windows Level Zero SDK path detection via LEVEL_ZERO_V1_SDK_PATH in CMakeLists.txt for building with Level Zero on Windows.
›Level Zero allocation path automatically disables itself with a warning at runtime if the detected devices do not support the Level Zero backend, preserving compatibility on non-dGPU or CPU-only systems.
llama.cpp b9141 adds continue_final_message body flag for vLLM/transformers API compatibility in the server and WebUI.
└──▷ GET THIS VERSION
$ git clone --branch b9141 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:$ git checkout b9141
└──▷ TRY IT
Trigger assistant prefill via the vLLM-compatible flag instead of server-side prefill_assistant — useful when integrating llama.cpp into pipelines that already target the vLLM API.
›Adds continue_final_message request body flag to the server API; when set with add_generation_prompt: false, it triggers the assistant prefill code path — aliasing the existing prefill_assistant heuristic for vLLM/transformers API compatibility.
›Enforces mutual exclusion between continue_final_message: true and add_generation_prompt: true, returning HTTP 400 on conflict, matching vLLM/transformers behavior.
›WebUI 'Continue' button now sends continue_final_message and add_generation_prompt: false automatically, with a matching opt-in option on the chat service.
Phoenix v15.9.0 backfills TOOL spans for external tool returns in the agent wrapper.
└──▷ GET THIS VERSION
$ git clone --branch arize-phoenix-v15.9.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:$ git checkout arize-phoenix-v15.9.0
›Backfills TOOL spans for external tool returns in the agent wrapper, improving trace completeness for agent workflows.
Milvus v2.6.16 adds streaming node resource group isolation, a new REST config endpoint, SIMD-accelerated IN queries, and Go SDK struct-array support.
└──▷ GET THIS VERSION
$ git clone --branch v2.6.16 https://github.com/milvus-io/milvus.git
# already have the repo? check out this version:$ git checkout v2.6.16
›Adds $partial_update field to the proxy access log for Upsert requests, exposing both explicit and implicitly promoted partial-update flags.
›Introduces streaming node resource group isolation with a new RESTful config inspection endpoint, allowing replicas to be assigned strictly within their configured resource groups.
›Exposes Arrow IO thread pool capacity as a refreshable paramtable knob to relieve HIGH-pool stalls under heavy storage v2 read load.
›Bumps Go SDK to v2.6.4 with full struct-array support (vector sub-fields, EmbeddingList search, schema validation), gRPC authority configuration, and preserved default gRPC dial options when custom DialOptions are provided.
›Accelerates TermExpr IN evaluation with a SIMD (AVX2/AVX512) batch filter, significantly improving query performance for IN predicates.
+7 moreshow less
›Increases the default L0 compaction deltalog max count from 30 to 1000 to reduce compaction backlog under high-delete workloads.
›Adds fast-fail retry capping and delegator stall detection so proxy queries failover to a healthy QueryNode immediately instead of exhausting the full backoff budget on a dead node.
›Parallelizes text match index loading on QueryNode to speed up segment load for collections with text indexes.
›Bypasses Knowhere search-pool scheduling for vector iterators to reduce per-Next overhead in iterator-heavy group-by search paths.
›Reduces proxy tail latency and memory pressure during traffic storms by fast-failing Enqueue before TSO/ID allocation and using a non-blocking edge-triggered task notifier.
›Rewrites sync manager key lock dispatcher with per-key FIFO queues and semaphore backpressure for non-blocking submission and graceful shutdown.
›Allows simultaneous pchannel increase and cluster/topology changes in replication config validation.
THE AI TOOLCHAIN