{
 "schema_version": 1,
 "generated": "2026-10-20",
 "id": "promethe-us-awesome-jev--gate-agent-tool-calls-with-jev-policy",
 "repo": "Promethe-us/awesome-jev",
 "repo_url": "https://github.com/Promethe-us/awesome-jev",
 "repo_slug": "promethe-us-awesome-jev",
 "title": "Hold risky agent tool calls for review",
 "page_title": "How to hold risky agent tool calls for review with Awesome Jev",
 "summary": "Let Jev score tool calls, then have your own policy approve, deny, or send them to a person.",
 "who": "A platform engineer running an agent that calls external tools and needs a review step for uncertain or high-risk actions.",
 "problem": "An agent may call a tool with side effects, and a model-only check is not a permission system.",
 "outcome": "A tool-call gate where Jev gives an advisory signal and application code, not Jev, makes the final approve, deny, or escalate decision.",
 "steps": [
  "Read the 'Jev 在 Agent Loop 中的位置' section and the 'Jev 不是什么' list to confirm Jev is a decision layer, not a security boundary.",
  "Open the Vercel eve evaluation guide linked in the 接入方式速查 table and review its `auto` and `evaluate` policy options.",
  "Set a policy that sends uncertain or high-risk tool calls to a human, and enforces permissions in your own code.",
  "Test the gate with blocked, allowed, and borderline tool calls, and confirm that tool execution never depends on Jev output alone."
 ],
 "agent_prompt": "Using the Promethe-us/awesome-jev README, outline a tool-call gate for <my agent> that uses Jev only as an advisory signal and keeps permission checks in application code. Return the policy steps and one test case per outcome, then verify that no tool runs without your own permission check.",
 "agents": [
  "any"
 ],
 "difficulty": "Advanced",
 "time": "90 minutes",
 "basis": "readme",
 "caveat": "The README states Jev is not a security boundary and not a physical robot safety controller, so permissions must live in your application code.",
 "category": "Research",
 "tags": [
  "Research"
 ],
 "lang": "",
 "requirements": [
  "A TypeSafe AI API key from the official docs, if you call the API directly",
  "Python or Node.js, if you use the SDKs or framework adapters",
  "A labeled set of example inputs to test Jev decisions before relying on them"
 ],
 "risk": "none",
 "seen_in_the_wild": false,
 "mentioned_publicly": false,
 "risks": [
  {
   "tag": "credentials",
   "notice": "The gate calls a paid, key-based API, so keep the key server-side and scope it to the gating service only."
  }
 ],
 "best": false,
 "score": 0.8,
 "evidence": [],
 "url": "https://dedupedai.com/use-cases/promethe-us-awesome-jev/gate-agent-tool-calls-with-jev-policy/",
 "md": "https://dedupedai.com/use-cases/promethe-us-awesome-jev/gate-agent-tool-calls-with-jev-policy/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/promethe-us-awesome-jev--gate-agent-tool-calls-with-jev-policy.json",
 "api_version": "v1"
}