---
schema_version: "newruntime-topic-hub-v0.1"
type: "topic_hub"
slug: "tool-use"
title: "Tool USE - New Runtime"
description: "A New Runtime topic hub collecting signals, patterns, field notes, and public sources about tool use."
answer: ["Tool USE is tracked here as an evidence-linked topic, not as a static glossary entry.","The page connects raw observations to pattern hypotheses, longer analysis, and public sources.","Use it as the canonical landing page before drilling into individual records."]
search_intents: ["tool use","tool use AI agents","tool use software"]
status: "tracked"
last_updated: "2026-07-21"
counts: {"total":8,"signals":6,"patterns":0,"posts":1,"atlas":1,"sources":18}
routes: {"html":"https://newruntime.com/topics/tool-use/","markdown":"https://newruntime.com/topics/tool-use.md","json":"https://newruntime.com/topics/tool-use.json"}
top_sources: ["https://askoracle.sh/","https://birdclaw.sh/","https://blog.google/products-and-platforms/products/search/search-io-2026","https://cameronrwolfe.substack.com/p/agent-evals","https://careersatdoordash.com/blog/building-doordash-assistant-an-engineering-overview","https://discrawl.sh/","https://gitcrawl.sh/","https://github.com/pydantic/pydantic-ai","https://imsg.sh/","https://mcporter.sh/","https://openai.com/index/advancing-voice-intelligence-with-new-models-in-the-api","https://pydantic.dev/articles/pydantic-ai-capabilities"]
---

# Tool USE - New Runtime

A New Runtime topic hub collecting signals, patterns, field notes, and public sources about tool use.

## Short Answer

- Tool USE is tracked here as an evidence-linked topic, not as a static glossary entry.
- The page connects raw observations to pattern hypotheses, longer analysis, and public sources.
- Use it as the canonical landing page before drilling into individual records.

## Patterns


## Field Notes

- [Agent APIs Need Fewer Magic Tricks and More Facts](https://newruntime.com/posts/agent-api-design/): Agent-first APIs should return explicit fields, precise errors, raw facts, and traceable metadata so models can repair tool calls deterministically instead of guessing world state.

## Recent Raw Signals

- 2026-06-28: [Pydantic AI loads capabilities only when needed](https://newruntime.com/signals/pydantic-ai-capabilities-progressive-disclosure/)
- 2026-06-24: [DoorDash Assistant Builds the Cart Instead of Giving Advice](https://newruntime.com/signals/doordash-assistant-builds-the-cart-instead-of-giving-advice/)
- 2026-05-21: [Agent Evals Become a Discipline Separate from Model Evals](https://newruntime.com/signals/agent-evals-become-a-discipline-separate-from-model-evals/)
- 2026-05-20: [Google Search Adds Agentic Action to Discovery](https://newruntime.com/signals/google-search-adds-agentic-action-to-discovery/)
- 2026-05-09: [A Personal Agent OS Emerges from Small CLI Tools](https://newruntime.com/signals/a-personal-agent-os-emerges-from-small-cli-tools/)
- 2026-05-09: [Voice Agents Move from Conversation to Tool-Using Work](https://newruntime.com/signals/voice-agents-move-from-conversation-to-tool-using-work/)

## Public Sources

- https://askoracle.sh/
- https://birdclaw.sh/
- https://blog.google/products-and-platforms/products/search/search-io-2026
- https://cameronrwolfe.substack.com/p/agent-evals
- https://careersatdoordash.com/blog/building-doordash-assistant-an-engineering-overview
- https://discrawl.sh/
- https://gitcrawl.sh/
- https://github.com/pydantic/pydantic-ai
- https://imsg.sh/
- https://mcporter.sh/
- https://openai.com/index/advancing-voice-intelligence-with-new-models-in-the-api
- https://pydantic.dev/articles/pydantic-ai-capabilities