{
 "schema_version": 1,
 "generated": "2026-10-20",
 "id": "w3cj-jev-chat--trace-why-a-tool-was-picked",
 "repo": "w3cj/jev-chat",
 "repo_url": "https://github.com/w3cj/jev-chat",
 "repo_slug": "w3cj-jev-chat",
 "title": "Trace why the chat bar picked a tool",
 "page_title": "How to trace why the chat bar picked a tool with Jev Chat",
 "summary": "Use the inspector to read every classifier answer and probability behind a reply.",
 "who": "A developer tuning or debugging tool routing who needs to know why a message went to one tool and not another.",
 "problem": "When a routed reply is wrong, it is hard to tell whether the classifier, the spell check or the policy caused it.",
 "outcome": "A clear record for any reply showing the request, each question, each probability and the action the code took.",
 "steps": [
  "Run `pnpm dev` and open http://localhost:5173.",
  "Send an ambiguous message, such as a short request that could fit two tools.",
  "Open the inspector pane for that reply.",
  "Check the original message, any spell or follow-up changes, and the probabilities for each tool choice.",
  "Run `pnpm test` to confirm the unit tests still pass after any change."
 ],
 "agent_prompt": "In the jev-chat repo, run pnpm dev, send the message <test-message> to the chat at http://localhost:5173, and open the inspector for the reply. Report the tool choice probabilities and the policy action. Verify that the inspector shows the original message and the answer to each question.",
 "agents": [
  "any"
 ],
 "difficulty": "Intermediate",
 "time": "30 minutes",
 "basis": "readme",
 "caveat": "The inspector shows what Jev answered and what the code did with it, but it does not change how Jev classifies a message.",
 "category": "Personal productivity",
 "tags": [
  "Personal productivity"
 ],
 "lang": "typescript",
 "requirements": [
  "Node 24 or later and pnpm 11 (run corepack enable to pick up the pinned version)",
  "A TypeSafe API key in TYPESAFE_API_KEY, since Jev is a paid API and the app shows error replies without it",
  "A terminal and a local browser on the same machine, because the API only listens on 127.0.0.1"
 ],
 "risk": "none",
 "seen_in_the_wild": false,
 "mentioned_publicly": false,
 "risks": [],
 "best": false,
 "score": 0.7,
 "evidence": [],
 "url": "https://dedupedai.com/use-cases/w3cj-jev-chat/trace-why-a-tool-was-picked/",
 "md": "https://dedupedai.com/use-cases/w3cj-jev-chat/trace-why-a-tool-was-picked/index.md",
 "disclaimer": "Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked \"Seen in the wild\", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email hello@dedupedai.com and we will look at it promptly.",
 "api": "https://dedupedai.com/api/v1/usecases/w3cj-jev-chat--trace-why-a-tool-was-picked.json",
 "api_version": "v1"
}