Use negative examples in skill descriptions for better routing
Add 'Don't call this skill when...' cases to skill descriptions to improve model routing accuracy.
Skill Description: Use when: User asks to summarize a document and provides the document text. Don't use when: User asks for a general knowledge question or asks to summarize a webpage link.
A practical pattern is to include a short “Use when vs. don’t use when” block directly in the description, and keep it concrete (inputs, tools involved, expected artifacts). A surprising failure mode is that making skills available can initially reduce correct triggering. One fix we’ve seen work is negative examples plus edge case coverage. In practice, that means writing a few explicit “Don’t call this skill when…” cases (and what to do instead). This helps the model route more cleanly, especially when you have multiple skills that look similar at a glance.
When you have multiple skills that might be confused, or when initial skill triggering is lower than expected.
- Skills might not be triggered correctly if descriptions are too vague or lack negative examples.
- Model may choose the wrong skill if multiple skills have overlapping functionality without clear distinctions.
Improve agent skill routing by adding negative examples and edge case coverage directly within skill descriptions. This clarifies when a skill should not be used, helping the model make cleaner routing decisions, especially with similar-looking skills.
Explicitly stating when *not* to use a skill provides clearer boundaries for the model, reducing ambiguity and improving the accuracy of skill selection, especially when skills have overlapping capabilities.
- Shell + Skills + Compaction: Tips for long-running agents that do real work ↗OpenAI Developers Blog · OpenAI Developers Blog · Sep 3, 2026 · quoted as shared
First seen on Subvolts Sep 3, 2026. Updated Sep 11, 2026. Spotted a problem? Tell us.
- Shell + Skills + Compaction: Tips for long-running agents that do real workOpenAI · OpenAI Developers Blog · Sep 3, 2026
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- ChecklistAX Practitioner Playbook Evaluation Quality ChecklistReview your AI agent evaluation process before running it to ensure its quality and trustworthiness. · 30 min
- Skill or agent configJoplin agent for automated note filing and retrievalAutomate note organization and retrieval in Joplin using a custom agent and Python scripts. · 60 min
- HeuristicUse reasoning models for improved reliability on complex tasksEmploy reasoning models like o1, o3, and o4-mini for dramatically improved reliability on complex, multi-step work. · 5 min
- Setting tipStore AI agent traces and evaluation data with projectKeep traces, evaluation datasets, and results with your project for repeatable testing. · 5 min
- WorkflowBuild and test an AI agent using Microsoft Foundry ToolkitCreate, configure, test, and evaluate an AI agent within VS Code. · 60 min
Questions people ask
- Which AI tools does this work with?
- Any current model. It was seen with in the sources.
- Where did this come from?
- OpenAI Developers Blog (OpenAI Developers Blog). Subvolts collected, normalized and credited it; the link above opens the original.