DEV Community

#llm

Posts

đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.
We fixed the worst prompt variant. It got better. That doesn't mean the fix worked.

We fixed the worst prompt variant. It got better. That doesn't mean the fix worked.

Comments 2
9 min read
Evals for Agents: Scoring Task Success, Trajectory, and Human Review

Evals for Agents: Scoring Task Success, Trajectory, and Human Review

Comments
8 min read
Per-Agent Cost and Token Attribution Straight From Your Traces

Per-Agent Cost and Token Attribution Straight From Your Traces

Comments
7 min read
The 2 a.m. Runaway Loop: Detecting and Killing an Agent That Won't Stop

The 2 a.m. Runaway Loop: Detecting and Killing an Agent That Won't Stop

Comments
7 min read
Agent State Desync: Why Your Agent Forgets and How to Fix It

Agent State Desync: Why Your Agent Forgets and How to Fix It

Comments
6 min read
Incident Response When Your Agent Is on Fire: A Runbook

Incident Response When Your Agent Is on Fire: A Runbook

Comments
6 min read
Context Bloat in Long-Running Agents: What to Keep, Summarize, and Drop

Context Bloat in Long-Running Agents: What to Keep, Summarize, and Drop

Comments
6 min read
The Agent Production-Readiness Checklist You Can Run Before Shipping

The Agent Production-Readiness Checklist You Can Run Before Shipping

Comments
7 min read
Planning and Decomposition for Agents: Structured Output Over Free-Form Reasoning

Planning and Decomposition for Agents: Structured Output Over Free-Form Reasoning

Comments
6 min read
The ReAct Loop From Scratch: The Agent Baseline Before Any Framework

The ReAct Loop From Scratch: The Agent Baseline Before Any Framework

Comments
7 min read
Agent Memory Done Right: Short-Term, Long-Term, and the Scratchpad

Agent Memory Done Right: Short-Term, Long-Term, and the Scratchpad

Comments
6 min read
Your First Tool-Calling Agent With No Framework, Just the Bare SDK

Your First Tool-Calling Agent With No Framework, Just the Bare SDK

Comments
6 min read
Meta's Agents Rule of Two: A Practical Defense Against Prompt Injection

Meta's Agents Rule of Two: A Practical Defense Against Prompt Injection

Comments
6 min read
Designing Agent Tools: Schemas and Descriptions That Prevent Misfires

Designing Agent Tools: Schemas and Descriptions That Prevent Misfires

Comments
6 min read
Choosing the Right Scraping Interface for LLM Workflows

Choosing the Right Scraping Interface for LLM Workflows

1
Comments
4 min read
đź‘‹ Sign in for the ability to sort posts by relevant, latest, or top.