The AI Toolchain — issue -342, September 8, 2025

Agno v1.8.2 adds response_model structured output support for AG-UI and Discord, plus updated_at on session responses.

$ git clone --branch v1.8.2 https://github.com/agno-agi/agno.git
# already have the repo? check out this version:
$ git checkout v1.8.2

from agno.client.discord import DiscordClient
from pydantic import BaseModel

class AnalysisResult(BaseModel):
    summary: str
    risk_level: str

client = DiscordClient(agent=my_agent, response_model=AnalysisResult)

• ›Adds response_model support to AG-UI, enabling structured output from agents in AG-UI apps.
• ›Adds response_model support to DiscordClient, enabling structured output from agents in Discord apps.
• ›Adds updated_at field to AgentSessionResponse and TeamSessionResponse.
• ›Extends JSON Schema constraint support for Gemini models.

LlamaIndex v0.14.0 adds document block support in OpenAI chat completions and upgrades to llama-index-workflows 2.0.

$ git clone --branch v0.14.0 https://github.com/run-llama/llama_index.git
# already have the repo? check out this version:
$ git checkout v0.14.0

• ›Adds support for document blocks in llama-index-llms-openai OpenAI chat completions.

browser-use 0.7.5 adds Qwen and Gemma LLM support, cross-origin iframe traversal, and exposes reasoning_models and cross_origin_iframes as parameters.

$ git clone --branch 0.7.5 https://github.com/browser-use/browser-use.git
# already have the repo? check out this version:
$ git checkout 0.7.5

agent = Agent(
    task='Book a flight on the airline website',
    llm=llm,
    reasoning_models=['qwen3']
)

• ›Exposes cross_origin_iframes as a parameter with depth limits, enabling the agent to traverse and interact with cross-origin iframes during browser automation.
• ›Exposes reasoning_models as a configurable parameter, allowing callers to explicitly declare which models use reasoning-style prompting.
• ›Adds Qwen LLM integration, expanding supported model providers.
• ›Adds Gemma LLM integration, expanding supported model providers.
• ›Adds paint-order filtering to improve accuracy of visible-element detection on complex pages.

Lightweight coding agent that runs in your terminal

Codex CLI 0.31.0 adds per-MCP-server startup timeouts, command preview highlighting, and richer session/status info.

$ git clone --branch rust-v0.31.0 https://github.com/openai/codex.git
# already have the repo? check out this version:
$ git checkout rust-v0.31.0

# In your codex config (e.g., ~/.codex/config.toml)
[mcp_servers.my_server]
command = "my-mcp-server"
startup_timeout_ms = 15000

• ›Adds startup_timeout_ms config option per MCP server declaration, fixing slow-start issues especially on Windows.
• ›Shows minutes and hours in the TUI thinking timer for long-running sessions.
• ›Displays CLI version in /status output.
• ›Respects model family configuration overrides from the TOML config file.
• ›Moves token usage and context information to the session level in the TUI.

The open source coding agent.

OpenCode v0.6.6 enables npm package installation on Windows and re-enables todo tools for Qwen models.

$ git clone --branch v0.6.6 https://github.com/sst/opencode.git
# already have the repo? check out this version:
$ git checkout v0.6.6

• ›Supports npm package installation on Windows.
• ›Re-enables todowrite and todoread tools for Qwen models.

An open-source AI agent that brings the power of Gemini directly into your terminal.

gemini-cli v0.5.0-nightly gains improved MCP slash command discoverability and a new UIContext hook.

$ git clone --branch v0.5.0-nightly.20250908.4693137b https://github.com/google-gemini/gemini-cli.git
# already have the repo? check out this version:
$ git checkout v0.5.0-nightly.20250908.4693137b

• ›Improves discoverability of MCP slash commands in the UI, making MCP integrations easier to find and use.
• ›Introduces useUI hook and UIContext for richer programmatic UI interaction.
• ›Allows slash commands to opt out of autocompletion and /help discovery for cleaner command surfaces.