LangSmith Trace Analyzer
**Role:** Applied AI engineer who reads LangSmith traces for a living. You've debugged 200+ production agent runs. **Context:** A specific …
Prompt
**Role:** Applied AI engineer who reads LangSmith traces for a living. You've debugged 200+ production agent runs. **Context:** A specific failed agent run trace: [PASTE OR DESCRIBE]. Expected behavior: [DESCRIBE]. Observed behavior: [DESCRIBE]. **Task:** Walk through the trace: 1. Identify the exact step where behavior diverged from intent. 2. Distinguish a model failure (model picked wrong action), a tool failure (tool returned bad data), a prompt failure (system prompt was ambiguous), or an architecture failure (loop didn't have a halt). 3. Pinpoint the root cause vs contributing factors. 4. Cite the exact line of system prompt or tool definition that needs to change. 5. Propose the specific fix. 6. Design the test that would prove the fix. 7. Identify if this is a one-off or a class of bugs. 8. Recommend the regression test to add. **Constraints:** - Never speculate without grounding in trace lines. - Distinguish "model was wrong" from "we asked the wrong question." - The fix must be falsifiable (a test that fails today and passes after). **Output format:** Trace walkthrough + diagnosis + fix + test.
How to use this prompt
- 1
Copy the prompt above and paste it into ChatGPT, Claude, or Gemini — or open it in the visual Studio to edit each part on a canvas and run it with your own key.
- 2
Replace any bracketed placeholders with your specifics. The more concrete your context and constraints, the sharper the result — see the 5-part prompt structure.
- 3
Run it, then refine. Ask the model to critique and improve its own answer with self-critique prompting.
Techniques in this prompt
Assigns the model an expert persona so it adopts the right vocabulary, depth, and standards for the task.
Learn this techniqueAsks the model to reason step by step before answering — ideal for multi-step, logical, or analytical tasks.
Learn this techniqueRecommended models
Build on this prompt
Open it in the visual Studio to wire it into a full workflow with your own API key — or learn the craft behind prompts like this.
More in AI Engineering
RAG vs Fine-tune Decision Memo
**Role:** You are a senior AI engineer who has shipped both RAG-based and fine-tuned LLM products at production scale. You believe most team…
Evals Harness Design for [Domain]
**Role:** AI engineer who has built evals suites that have caught 30+ production regressions before they shipped. You believe vibes-based "t…
System Prompt Audit
**Role:** Senior prompt engineer who has audited 100+ production system prompts. You read prompts the way an editor reads prose — for the me…
Agent Loop Halt-Condition Design
**Role:** Applied AI engineer who has shipped agents that completed millions of tool-calling iterations in production. You believe most agen…