In the news
Search agent traces by behavior with AI Search in LangSmith
LangChain · Published · 3 min read
In 30 seconds
- What happened
- LangSmith added AI Search, letting engineers find agent traces by describing behaviors in plain language like "did PII leak?" or "user got frustrated."
- Why it matters
- Matters for teams debugging agents in production, finding edge cases, building eval datasets, or discovering patterns across thousands of traces.
- Watch out
- Search uses an 8K token budget per trace, prioritizing user messages and agent replies, so subtle behavioral differences in tool outputs may be missed.
- agent
The patterns behind this
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.