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 -345, September 5, 2025

THE AI TOOLCHAIN NO. -345
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED SEPTEMBER 5, 2025 · 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   # 6 tools matched
AI & LLM Tooling
◆  AI Coding Agents

Charm Crush

Sources Release notes → v0.7.5 NOTES

Glamourous agentic coding for all

Crush v0.7.5 shows the persistent shell path in the permission dialog for better shell-action transparency.

└──▷ GET THIS VERSION
$ git clone --branch v0.7.5 https://github.com/charmbracelet/crush.git
# already have the repo? check out this version:
$ git checkout v0.7.5
  • Displays the persistent shell path in the permission dialog so users can see exactly which shell session is being granted access.
Was this useful?

OpenAI Codex CLI

Sources Release notes → rust-v0.30.0 NOTES

Lightweight coding agent that runs in your terminal

Codex CLI 0.30.0 adds session resume/continue flags, MCP session history, bash syntax highlighting, and Windows Full Access auto-approval.

└──▷ GET THIS VERSION
$ git clone --branch rust-v0.30.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.30.0
└──▷ TRY IT
Return to a previous Codex session interactively — useful for long-running tasks you stepped away from.
$ codex --resume
Instantly re-enter the most recent session without navigating a picker, ideal for quick follow-up prompts.
$ codex --continue
  • Adds --resume session picker and --continue quick-resume flags to the TUI for returning to previous sessions.
  • Adds MCP session resume and history listing support, replaying event messages from response items on resume.
  • Adds bash syntax highlighting in the TUI command display.
  • Enables auto-approval of patches on Windows in Full Access (DangerFullAccess) mode, eliminating approval prompts on non-sandboxed platforms.
  • Improves @ file search to include specific hidden directories such as .github and .gitlab.
+6 moreshow less
  • Introduces a rollout policy mechanism for controlled feature exposure.
  • Adds a shared HTTP client for more consistent network handling.
  • Uses macOS-native key glyphs (⌥⇧⌃) for keyboard hints on Mac.
  • Pager now auto-scrolls (pins scroll to bottom) in the TUI.
  • Introduces MCP sandbox call support.
  • Prompts Codex to read AGENTS.md files for project-level agent instructions.
└──▷ BREAKING ON UPGRADE
  • !Project .env files are no longer loaded automatically — environment variables previously sourced from a project .env on startup must now be set in the shell or passed explicitly.
Was this useful?
◆  AI Agent Frameworks

PydanticAI

Sources Release notes → v1.0.0 NOTES

PydanticAI v1.0.0 adds human-in-the-loop tool approval, LiteLLM provider, tool-call usage limits, and NativeOutput for Groq.

└──▷ GET THIS VERSION
$ git clone --branch v1.0.0 https://github.com/pydantic/pydantic-ai.git
# already have the repo? check out this version:
$ git checkout v1.0.0
└──▷ USE IT
Cap the number of tool calls an agent can make in a single run to prevent runaway tool loops.
python
from pydantic_ai import Agent
from pydantic_ai.settings import UsageLimits

agent = Agent('openai:gpt-4o')
result = await agent.run(
    'Research and summarize the latest CVEs for OpenSSL',
    usage_limits=UsageLimits(tool_calls_limit=5)
)
print(result.usage().tool_calls)  # inspect actual tool calls made
Control how tool schemas are generated from docstrings — useful when enforcing that all parameters must be documented before deployment.
python
from pydantic_ai.toolsets import FunctionToolset

toolset = FunctionToolset(
    docstring_format='google',
    require_parameter_descriptions=True
)
  • Adds tool_calls_limit to UsageLimits and tool_calls to RunUsage to cap and track tool-call counts per run.
  • Adds docstring_format, require_parameter_descriptions, and schema_generator parameters to FunctionToolset for fine-grained tool schema control.
  • Adds operation.cost span attribute to model request spans; renames ModelResponse.price() to ModelResponse.cost().
  • Adds identifier field to FileUrl and its subclasses.
  • Adds human-in-the-loop tool call approval support, enabling agents to pause and await human confirmation before executing tools.
+5 moreshow less
  • Adds a LiteLLM provider for OpenAI-API-compatible models.
  • Supports NativeOutput with Groq models.
  • Bundles logfire with the pydantic-ai package so tracing is available without a separate install.
  • Allows most types used in documentation examples to be imported directly from pydantic_ai.
  • Defaults InstrumentationSettings version to 2.
└──▷ BREAKING ON UPGRADE
  • !Python 3.9 is no longer supported; the minimum supported version is now Python 3.10.
  • !ModelResponse.price() is renamed to ModelResponse.cost(); call sites using .price() will break.
  • !OpenAIModelProfile.openai_supports_sampling_settings is deprecated.
  • !mcp-run-python has been moved to its own repository and is no longer part of this package.
Was this useful?

browser-use

Sources Release notes → 0.7.2 NOTES

browser-use 0.7.2 adds sample_images support and an AgentMail integration alongside hook API updates.

└──▷ GET THIS VERSION
$ git clone --branch 0.7.2 https://github.com/browser-use/browser-use.git
# already have the repo? check out this version:
$ git checkout 0.7.2
  • Adds sample_images parameter to let the agent learn how to operate a platform from example screenshots.
  • New AgentMail integration for browser automation workflows involving email.
  • Updated hooks APIs following Playwright removal, with revised documentation.
Was this useful?
Other / Uncategorized
◆  AI OBSERVABILITY

Langfuse

Sources Release notes → v3.108.0 NOTES

Langfuse v3.108.0 revamps in-app support with acknowledgement emails and adds a structured GitHub Discussions form.

└──▷ GET THIS VERSION
$ git clone --branch v3.108.0 https://github.com/langfuse/langfuse.git
# already have the repo? check out this version:
$ git checkout v3.108.0
  • Revamps the in-app support experience with a new interface and adds a structured form for GitHub Discussions.
  • Sends an acknowledgement email to users automatically when a new support thread is created.
  • Delivers styling improvements to agent graphs in the trace UI.
Was this useful?
◆  VECTOR DB RAG

Weaviate

Sources Release notes → v1.32.6 NOTES

Weaviate v1.32.6 adds Amazon Nova model support, a new multi2vec-aws module, a text2vec-morph module, and alias-based GET collection operations.

└──▷ GET THIS VERSION
$ git clone --branch v1.32.6 https://github.com/weaviate/weaviate.git
# already have the repo? check out this version:
$ git checkout v1.32.6
  • Adds multi2vec-aws module, enabling multimodal vectorization via AWS.
  • Adds text2vec-morph module for text vectorization via Morph.
  • Adds support for Amazon Nova models in the modules layer.
  • Adds maxTokens support in the generative-aws module.
  • Supports GET collection operations via alias, expanding alias-based API coverage.
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 →