Heads up This site is currently under heavy development.
Subscribe Get it delivered — the daily firehose, filtered to the tools you run, plus the documentation changes vendors never announce. Compare plans →

The AI Toolchain — issue -051, June 29, 2026

THE AI TOOLCHAIN NO. -051
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED JUNE 29, 2026 · EVERY WEEKDAY
EDITIONS tail grep head diff uniq

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.

VIEW
ISSUE VIEW full issue
Do you prefer this view?
$ tct list   # 8 tools matched
AI & LLM Tooling
◆  AI Coding Agents

Anthropic Claude Code

Sources Release notes → v2.1.196 NOTES

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 v2.1.196 adds org-default models, clickable file attachments, and an always-on stream watchdog.

└──▷ GET THIS VERSION
$ git clone --branch v2.1.196 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.196
└──▷ TRY IT
Disable the new always-on stream watchdog if your provider intentionally sends long silent gaps (e.g. a slow self-hosted endpoint).
$ CLAUDE_ENABLE_STREAM_WATCHDOG=0 claude
  • Adds organization default models: admins set a model in the org console; users see 'Org default' or 'Role default' in /model when no personal selection is made.
  • Adds readable default session names at start, making sessions easier to identify and resume.
  • Adds clickable file attachments in chat — Cmd/Ctrl-click reveals the file in Finder/Explorer.
  • Enables the streaming idle watchdog by default for all providers: aborts and retries any response stream that produces no events for 5 minutes; disable with CLAUDE_ENABLE_STREAM_WATCHDOG=0.
  • Improves background agents: workers killed by a daemon restart are automatically resumed from where they left off when the agents view next opens.
+3 moreshow less
  • Improves /code-review workflow: merges five cleanup finders into one, cutting token usage by roughly 25%.
  • Disables Remote Control automatically when ANTHROPIC_BASE_URL points at a non-Anthropic host, matching existing behavior under CLAUDE_CODE_USE_BEDROCK/_VERTEX/_FOUNDRY.
  • Hardens claude mcp list/get: untrusted workspaces no longer auto-spawn .mcp.json servers self-approved via a committed .claude/settings.json; such servers show ⏸ Pending approval instead.
└──▷ BREAKING ON UPGRADE
  • !Opening the agents view from a foreground session now requires a single press instead of two.
Was this useful?

Cline

Sources Release notes → v4.0.4 3 RELEASES · 2026-06-29 NOTES STABLE

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

Cline v4.0.4 makes ClinePass available everywhere in the UI — onboarding, settings, and credit-limit actions.

└──▷ GET THIS VERSION
$ git clone --branch v4.0.4 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v4.0.4
  • ClinePass is now universally available across all UI surfaces: onboarding, settings, the welcome promo banner, and the credit-limit 'Switch to ClinePass' action.
2 more releases in this issue · 2026-06-29
v4.0.3 NOTES STABLE

ClinePass provider is now available to all users, no longer behind a feature flag.

└──▷ GET THIS VERSION
$ git clone --branch v4.0.3 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v4.0.3
  • Enables the ClinePass provider for all users by removing the feature-flag gate.
v4.0.2 NOTES STABLE

Cline v4.0.2 adds xhigh reasoning effort support for DeepSeek thinking models.

└──▷ GET THIS VERSION
$ git clone --branch v4.0.2 https://github.com/cline/cline.git
# already have the repo? check out this version:
$ git checkout v4.0.2
  • Adds reasoning effort support, including the xhigh level, for DeepSeek thinking models.
  • Improves ClinePass provider experience with clearer reasoning controls and model selection UI.
Was this useful?

Charm Crush

Sources Release notes → v0.81.0 NOTES

Glamourous agentic coding for all

Crush v0.81.0 adds a configurable chat scrollbar, herdr.dev agent status integration, and expandable tool-call names.

└──▷ GET THIS VERSION
$ git clone --branch v0.81.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.81.0
└──▷ USE IT
Keep the scrollbar always visible during long sessions instead of auto-hiding after 2 seconds.
json
{
  "options": {
    "tui": {
      "scrollbar": "always"
    }
  }
}
  • Adds a scrollbar to the chat view, configurable via options.tui.scrollbar (default, always, or never).
  • Integrates with herdr.dev via a herdr socket, enabling agent status signaling (blocked/finished notifications).
  • Enables expanding tool-call names in the UI for easier inspection of agent activity.
Was this useful?
◆  AI Agent Frameworks

PydanticAI

Sources Release notes → v2.1.0 NOTES

PydanticAI v2.1.0 adds Anthropic web tools with server-tool replay, TypeAdapter for EvaluatorContext, and improved instrumentation serialization.

└──▷ GET THIS VERSION
$ git clone --branch v2.1.0 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v2.1.0
  • Adds TypeAdapter support for EvaluatorContext, enabling structured validation and serialization of evaluator context objects.
  • Adds Anthropic _20260209 web tools with server-tool replay support for the Anthropic provider.
  • Serializes instrumentation message attributes using to_json instead of json.dumps for more robust OpenTelemetry attribute handling.
Was this useful?
◆  Local LLM Runtimes

llama.cpp

Sources Release notes → b9840 2 RELEASES · 2026-06-29 NOTES STABLE

llama.cpp b9840 adds DeepSeek V4 model support including conversion, multi-sequence inference, and Flash Attention.

└──▷ GET THIS VERSION
$ git clone --branch b9840 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b9840
  • Adds dsv4 architecture support to the convert tooling, enabling conversion of DeepSeek V4 (and Pro) models to GGUF format.
  • Supports DeepSeek V4 inference with Flash Attention (FA) enabled, graph reuse, multi-sequence support, and save/load state.
  • Adds compatibility with antirez-format DeepSeek V4 GGUFs, broadening which existing GGUF files can be loaded.
  • Uses expert_gating_func in place of moe.score_func for DeepSeek V4 MoE gating configuration.
1 more release in this issue · 2026-06-29
b9837 NOTES STABLE

llama.cpp b9837 adds --reasoning-preserve flag for controlling reasoning content in chat output.

└──▷ GET THIS VERSION
$ git clone --branch b9837 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b9837
  • Adds --reasoning-preserve flag to control whether reasoning/chain-of-thought content is preserved in chat and Jinja template output.
Was this useful?

vLLM

Sources Release notes → v0.24.0 NOTES

vLLM v0.24.0 adds MiniMax-M3, DiffusionGemma, a Streaming Parser Engine, Rust frontend endpoints, and removes internal CUDA_VISIBLE_DEVICES management.

└──▷ GET THIS VERSION
$ git clone --branch v0.24.0 https://github.com/vllm-project/vllm.git
# already have the repo? check out this version:
$ git checkout v0.24.0
└──▷ TRY IT
Skip Triton autotuning at startup to reduce cold-start time in latency-sensitive deployments.
$ VLLM_TRITON_FORCE_FIRST_CONFIG=1 vllm serve deepseek-ai/DeepSeek-V3 --tensor-parallel-size 4
Pause the Rust frontend to drain in-flight requests during a rolling restart, then resume once the new instance is ready.
$ curl -X POST http://localhost:8000/pause && sleep 5 && curl -X POST http://localhost:8000/resume
  • Adds device_ids argument to replace internal CUDA_VISIBLE_DEVICES management; vLLM no longer sets CUDA_VISIBLE_DEVICES internally.
  • Adds VLLM_TRITON_FORCE_FIRST_CONFIG environment variable to skip Triton autotuning and cut startup overhead.
  • Rust frontend gains API-key authentication, CORS, /tokenize, /detokenize, /pause, /resume, /is_paused, /abort_requests, /get_world_size, thinking_token_budget, and parallel_tool_calls=false.
  • Adds real /v1/embeddings support for messages with chat_template_kwargs.
  • Adds cache-usage reporting to the Anthropic-compatible /v1/messages endpoint.
+29 moreshow less
  • Adds multimodal token counts inside usage.prompt_tokens_details in API responses.
  • Adds strict mode for tool calling in Chat Completions API and Responses API.
  • Introduces a new Streaming Parser Engine that unifies tool-call and reasoning parsing, with parsers for Qwen3, MiniMax-M2, GLM-4.7/5.1/5.2, and Nemotron V3.
  • Adds support for new models: MiniMax-M3, DiffusionGemma (including a CPU path), Hierarchical Reasoning Model (HrmTextForCausalLM), and OpenMOSS.
  • Model Runner V2 now supports quantized models by default and enables GraniteMoE by default.
  • Adds Dynamic Speculative Decoding and EAGLE3 support for Qwen3 targets.
  • Adds online FP8 per-token-per-channel (PTPC) quantization.
  • Extends modelopt_mixed quantization support to Ampere (SM80-86) and Turing (SM75) GPUs.
  • Allows fp8_e5m2 KV cache for non-FP8 checkpoints.
  • Adds flashinfer_cutlass as a clamped NVFP4 MoE backend.
  • Integrates DeepEP v2 for expert parallelism.
  • Adds KV push from prefill to decode via NIXL for disaggregated serving.
  • Adds a KV-cache watermark to reduce preemptions.
  • Adds a Marconi-style admission policy for hybrid KV cache.
  • Adds structured-output guardrails for diffusion decoders.
  • Adds structured outputs support for beam search.
  • Adds Qwen3-VL video loader and Qwen2-VL/Qwen2.5-VL processor-mapped video loader.
  • Adds ViT full CUDA graph support for GLM-4.1V, DeepSeek-OCR, Kimi-VL, mllama4, and Lfm2VL encoder.
  • Adds fastsafetensors ParallelLoader for faster weight loading.
  • Adds 2.5x faster ASR CPU preprocessing via multi-threading.
  • Adds cgroup memory-limit-aware KV cache sizing on CPU.
  • Adds RISC-V oneDNN W8A8 INT8 support and RVV micro-GEMM for WNA16.
  • Migrates GGUF quantization to a plugin.
  • Adds Intel XPU sequence-parallel support and W4A16 int4 group_size=32 MoE.
  • Adds SM90 CUTLASS FP8 odd-M support via swap_ab (180-290% kernel speedup).
  • Adds FlashInfer cutedsl NVFP4 GEMM backend and cute-dsl MXFP8 linear kernel.
  • Adds Helion kernels for FP8/RMSNorm quantization.
  • Completes migration of all kernels to the libtorch stable ABI.
  • Deprecation window begun for CUDA_VISIBLE_DEVICES on ROCm.
└──▷ BREAKING ON UPGRADE
  • !ROCm: CUDA_VISIBLE_DEVICES support has entered a deprecation window and will be removed in a future release.
  • !P2pNcclConnector has been removed; configurations using it must migrate to an alternative KV connector.
Was this useful?
Other / Uncategorized
◆  VECTOR DB RAG

LanceDB

Sources Release notes → v0.31.0-beta.4 2 RELEASES · 2026-06-29 NOTES STABLE

LanceDB v0.31.0-beta.4 adds an OAuth header provider for Rust clients.

└──▷ GET THIS VERSION
$ git clone --branch v0.31.0-beta.4 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout v0.31.0-beta.4
  • Adds an OAuth header provider to the Rust client, enabling authenticated requests to LanceDB services via OAuth.
1 more release in this issue · 2026-06-29
python-v0.34.0-beta.4 NOTES STABLE

LanceDB python-v0.34.0-beta.4 adds an OAuth header provider for authenticated connections.

└──▷ GET THIS VERSION
$ git clone --branch python-v0.34.0-beta.4 https://github.com/lancedb/lancedb.git
# already have the repo? check out this version:
$ git checkout python-v0.34.0-beta.4
  • Adds an OAuth header provider for authenticating requests, enabling token-based auth flows when connecting to LanceDB services.
Was this useful?
◆  AI OBSERVABILITY

Langfuse

Sources Release notes → v3.202.0 NOTES

Langfuse v3.202.0 adds a generic short-lived credential refresh manager and conversation deletion in the agent UI.

└──▷ GET THIS VERSION
$ git clone --branch v3.202.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.202.0
  • Adds a generic short-lived credential interface and refresh manager for rotating ephemeral credentials automatically.
  • Enables deleting conversations in the agent interface.
Was this useful?
my-toolchain — 0 tools
paste an install list to detect your tools

A brew list, a Brewfile, requirements.txt, a Dockerfile — or just the product names, free-form. Nothing leaves your browser.

    browse all tools →