AG-UI news
Every thread that mentions this topic, across all sections
Moka Chat gives Ollama models MCP tools via one npx command, with an inspector showing every tool call
A developer released Moka, a free, MIT-licensed local sandbox for testing prompts and MCP servers visually. It runs via 'npx @mokalabs/sandbox' with zero config, auto-detects running Ollama instances, and includes an inspector exposing raw JSON-RPC traffic, tokens, and latency. Developers can test tool calls and streaming, compare models side by side, and export working LangGraph Python scripts directly.
A developer released Moka, a free, MIT-licensed local sandbox for testing prompts and MCP servers visually. It runs via 'npx @mokalabs/sandbox' with zero config, auto-detects running Ollama instances, and includes an inspector exposing raw JSON-RPC traffic, tokens, and latency. Developers can test tool calls and streaming, compare models side by side, and export working LangGraph Python scripts directly.
Show HN: Plurnk, an open-source AI coding harness with model-curated context and 200MB footprint
A developer released Plurnk, an open-source AI harness dogfooded for weeks as a daily driver on an RTX 5070 Ti with 16GB VRAM. It emphasizes model-managed context curation, an ANTLR grammar for its verbs, and universal resource addressing via pseudo-URIs. The author seeks feedback and targets builders running agentic workflows under local, modest hardware constraints.
A developer released Plurnk, an open-source AI harness dogfooded for weeks as a daily driver on an RTX 5070 Ti with 16GB VRAM. It emphasizes model-managed context curation, an ANTLR grammar for its verbs, and universal resource addressing via pseudo-URIs. The author seeks feedback and targets builders running agentic workflows under local, modest hardware constraints.
pydantic-ai v2.37.0 adds glm-5.3-flash support and Z.AI fixes
Pydantic AI released v2.37.0 (2026-08-31), adding the glm-5.3-flash model and reworking the Z.AI test suite onto cassettes. The release also fixes Z.AI's non-standard finish_reason mapping and AG-UI streaming behavior, improving agent-framework compatibility with Z.AI models.
Pydantic AI released v2.37.0 (2026-08-31), adding the glm-5.3-flash model and reworking the Z.AI test suite onto cassettes. The release also fixes Z.AI's non-standard finish_reason mapping and AG-UI streaming behavior, improving agent-framework compatibility with Z.AI models.
Anthropic releases AG-UI adapter cookbook for Claude Managed Agents
Anthropic published a cookbook pairing Claude Managed Agents with AG-UI, CopilotKit's open Agent-User Interaction Protocol. The included adapter maps each chat thread to a managed session and streams text, tool calls, and thinking to custom UIs.
Anthropic published a cookbook pairing Claude Managed Agents with AG-UI, CopilotKit's open Agent-User Interaction Protocol. The included adapter maps each chat thread to a managed session and streams text, tool calls, and thinking to custom UIs.