Backfill issue -185 — The AI Toolchain

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

v2.1.41 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 v2.1.41 adds auth subcommands, Windows ARM64 support, and smarter /rename.

└──▷ GET THIS VERSION

$ git clone --branch v2.1.41 https://github.com/anthropics/claude-code.git
# already have the repo? check out this version:
$ git checkout v2.1.41

└──▷ TRY IT

Check your current authentication state without opening a browser — useful in CI or remote SSH sessions.

$ claude auth status

$ claude auth login

Let Claude name the current session automatically based on what you've discussed, instead of typing a name manually.

$ /rename

›Adds claude auth login, claude auth status, and claude auth logout CLI subcommands for managing authentication from the terminal.
›Adds native Windows ARM64 (win32-arm64) binary support.
›Improves /rename to auto-generate a session name from conversation context when called without arguments.

Was this useful?

Charm Crush

v0.43.0 NOTES

Glamourous agentic coding for all

Crush v0.43.0 adds a toggle todos/pills menu item and glob-based gitignore for faster monorepo handling.

└──▷ GET THIS VERSION

$ git clone --branch v0.43.0 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.43.0

›Adds a toggle todos/pills menu item to the UI.
›Replaces regex-based gitignore parsing with glob-based matching, reducing CPU and disk usage on large repositories.

Was this useful?

SST OpenCode

v1.1.65 NOTES

The open source coding agent.

OpenCode v1.1.65 adds a tool.definition plugin hook and desktop improvements including model toggling and auto-reconnect.

└──▷ GET THIS VERSION

$ git clone --branch v1.1.65 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v1.1.65

└──▷ USE IT

Override a built-in tool's description or parameters from a plugin to restrict or annotate its behavior for your team.

javascript

// In your opencode plugin
export default {
  hooks: {
    "tool.definition": (tool) => {
      if (tool.name === "bash") {
        tool.description = "Run shell commands (restricted: no sudo allowed)";
      }
      return tool;
    }
  }
}

›Adds tool.definition hook for plugins to modify tool descriptions and parameters at runtime.
›Enables toggling all provider models at once in desktop settings.
›Reconnects the event stream automatically on disconnect in the desktop client.
›Desktop notifications now navigate directly to the associated session.

└──▷ BREAKING ON UPGRADE

!Worktree delete functionality has been removed.

Was this useful?

Earendil Works Pi

v0.52.12 NOTES

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

Pi v0.52.12 adds a transport setting to choose between SSE, WebSocket, or auto for supported providers.

└──▷ GET THIS VERSION

$ git clone --branch v0.52.12 https://github.com/earendil-works/pi.git
# already have the repo? check out this version:
$ git checkout v0.52.12

└──▷ USE IT

Force SSE transport for openai-codex when WebSocket connections are blocked by a proxy.

json

# In settings.json
{
  "transport": "sse"
}

›New transport setting ("sse", "websocket", "auto") in /settings and settings.json lets you control the connection transport for providers like openai-codex.
›Interactive mode applies transport changes immediately without restarting the agent session.

└──▷ BREAKING ON UPGRADE

!The legacy websockets: boolean setting is migrated to the new transport setting; existing configs using websockets will be remapped automatically.

Was this useful?

◆ AI Agent Frameworks

OpenAI Agents SDK

v0.9.0 NOTES

OpenAI Agents SDK v0.9.0 adds configurable function-tool timeouts and a ToolOutputTrimmer for smart context management.

└──▷ GET THIS VERSION

$ git clone --branch v0.9.0 https://github.com/openai/openai-agents-python.git
# already have the repo? check out this version:
$ git checkout v0.9.0

└──▷ USE IT

Cap a slow function tool at 5 seconds and surface the timeout as a result string rather than crashing the run.

python

from agents import function_tool, ToolTimeoutBehavior

@function_tool(timeout_seconds=5.0, timeout_behavior="error_as_result")
def slow_lookup(query: str) -> str:
    ...  # long-running external call

›Adds timeout_seconds, timeout_behavior, and timeout_error_function parameters to function tools, letting you cap execution time and choose between 'error_as_result' or 'raise_exception' on timeout via ToolTimeoutBehavior and ToolErrorFunction.
›Adds ToolOutputTrimmer for smart context management, enabling automatic trimming of tool output to fit within context limits.

└──▷ BREAKING ON UPGRADE

!Python 3.9 is no longer supported; upgrade to Python 3.10 or newer.
!Agent.as_tool() now returns FunctionTool instead of the broader Tool union type; code that depends on the Tool return type may require adjustment.

Was this useful?

OpenClaw

v2026.2.12 NOTES

Your own personal AI assistant. Any OS.

OpenClaw v2026.2.12 adds local-time log display, Telegram blockquote rendering, and Z.AI/MiniMax provider expansions.

└──▷ GET THIS VERSION

$ git clone --branch v2026.2.12 https://github.com/openclaw/openclaw.git
# already have the repo? check out this version:
$ git checkout v2026.2.12

└──▷ TRY IT

Tail agent logs with timestamps converted to your local timezone so you can correlate events without mental UTC math.

$ openclaw logs --local-time

›Adds openclaw logs --local-time flag to display log timestamps in the local timezone with ±HH:MM offsets.
›Telegram: renders blockquotes as native <blockquote> tags instead of stripping them.
›Onboarding: adds Z.AI endpoint-specific auth choices (zai-coding-global, zai-coding-cn, zai-global, zai-cn) and expands default Z.AI model wiring.
›Onboarding: adds MiniMax M2.5 and M2.5-Lightning model entries, updates default/recommended models from M2.1 to M2.5, and includes minimax-m2.5 in modern model filtering.
›Discord: allows channel-edit to archive/lock threads and set auto-archive duration.

└──▷ BREAKING ON UPGRADE

!POST /hooks/agent now rejects payload sessionKey overrides by default; set hooks.defaultSessionKey (recommended with hooks.allowedSessionKeyPrefixes: ["hook:"]) to keep fixed hook context, or explicitly set hooks.allowRequestSessionKey: true to restore legacy behavior.

Was this useful?

◆ Local LLM Runtimes

llama.cpp

b8021 2 RELEASES · 2026-02-13 NOTES STABLE

llama.cpp b8021 adds OpenCL support for Q4_1 quantization in matrix-vector and matrix-matrix operations.

└──▷ GET THIS VERSION

$ git clone --branch b8021 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8021

›Adds OpenCL backend support for q4_1 quantized matrix-vector (mv) and matrix-matrix (mm) operations, enabling GPU-accelerated inference with Q4_1 models on OpenCL devices.
›Adds general q4_0 matrix-matrix (mm) support in the OpenCL backend.

1 more release in this issue · 2026-02-13

b8020 NOTES STABLE

llama-parallel gains the -kvu flag in b8020

└──▷ GET THIS VERSION

$ git clone --branch b8020 https://github.com/ggml-org/llama.cpp.git
# already have the repo? check out this version:
$ git checkout b8020

›Adds -kvu argument to llama-parallel for controlling key-value cache usage during parallel inference.

Was this useful?

Other / Uncategorized

◆ AI OBSERVABILITY

Arize Phoenix

arize-phoenix-evals-v2.10.0 3 RELEASES · 2026-02-13 NOTES STABLE

Phoenix Evals 2.10.0 adds dataset evaluators, a tool-response evaluator template, label filtering for classifiers, and full Mustache prompt support.

└──▷ GET THIS VERSION

$ git clone --branch arize-phoenix-evals-v2.10.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-evals-v2.10.0

›Adds dataset evaluators, enabling evaluation runs directly against datasets.
›Adds a tool response handling evaluator template for assessing LLM tool-call outputs.
›Adds label-based filtering for built-in classification evaluators, letting practitioners scope results to specific labels.
›Enables full Mustache prompt templates on the server side for more expressive eval prompt authoring.

2 more releases in this issue · 2026-02-13

arize-phoenix-client-v1.29.0 NOTES STABLE

Arize Phoenix client v1.29.0 adds dataset evaluators for running evaluations over datasets.

└──▷ GET THIS VERSION

$ git clone --branch arize-phoenix-client-v1.29.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-client-v1.29.0

›Adds dataset evaluators, enabling evaluation runs to be applied across datasets in the Phoenix client.

arize-phoenix-v13.0.0 NOTES STABLE

Phoenix v13 ships dataset evaluators, built-in LLM eval configs in GraphQL, OpenAI Responses API support, and async Bedrock client.

└──▷ GET THIS VERSION

$ git clone --branch arize-phoenix-v13.0.0 https://github.com/Arize-ai/phoenix.git
# already have the repo? check out this version:
$ git checkout arize-phoenix-v13.0.0

›Adds built-in LLM evaluator configs to the GraphQL API, exposing pre-built evaluator templates directly via the graph.
›Adds support for OpenAI API type selection between Chat Completions and Responses API in the playground.
›Converts the Bedrock client from boto3 to aioboto3 for fully async operation, unlocking non-blocking LLM calls.
›Adds a session ID index for spans across SQLite and Postgres backends, improving span query performance at scale.
›Introduces dataset evaluators — a new first-class object that attaches evaluators directly to datasets.

+16 moreshow less

›Adds pre-built LLM evaluators (built-ins) to the evaluator creation menu, with input mapping, output config display, config overrides, and a dedicated built-in evaluator table.
›Adds a tool response handling evaluator template for evaluating LLM tool call outputs.
›Adds a JSON parse toggle to the JSON distance built-in evaluator.
›Adds custom LLM provider support: create, edit, delete custom providers from the model menu, with secrets management.
›Adds model search to the model menu for faster provider/model selection.
›Adds an evaluator preview mutation to the GraphQL API for testing evaluators before saving.
›Adds an explanation toggle to the evaluator form to control whether evaluations include reasoning.
›Adds an append messages parameter in the playground for controlling message chaining.
›Adds available tools display to experiment output so tool-calling behavior is visible alongside results.
›Adds a dataset deep link after dataset selection from the playground.
›Adds an examples route with an examples table for browsing dataset examples directly.
›Enables creating dataset examples in a chain (sequential example creation flow).
›Deserializes JSON in message function/tool call arguments via dataset_helpers.
›Adds user ID tracking on evaluators for attribution.
›Adds eval outputs panel to the playground for inline evaluation results.
›Adds autocomplete to the LLM eval prompt editor.

Was this useful?

// END OF TRANSMISSION

The AI Toolchain · curated daily · February 13, 2026
Releases & features from the open-source and commercial AI tools that practitioners actually run.

The AI Toolchain — issue -185, February 13, 2026