Explore / Case study / Codex + Jev Batch Classify & Score
Codex + Jev Batch Classify & Score
Demo onlyTerminalWebInstalls a Jev skill in Codex to batch-run Choice/Score questions for content cleaning and agent go/no-go.
by ya piao · listed 5 days ago
⚡ What System1 does here
For each item in a batch, Jev answers typed Choice or Score questions (e.g. keep/drop, relevance, quality). Codex only acts on the structured labels; Jev does not write explanations or long prose.
About
Community case based on a public walkthrough: load an official/community Jev skill into Codex, then ask many classification or scoring questions in one pass for data cleaning and agent decision points. Evidence (third-party public post, not an official TypeSafe customer story): - https://x.com/NFT_Chen/status/2101644445078728802 Limits to keep clear: this is a developer demo / tutorial pattern, not a named production deployment. Jev is used for structured decisions; code or SQL should still do arithmetic and deterministic rules.
🧪 Community test reports
0 reports · 0 passed on the real modelNo one has tested this yet. Be the first.
💬 Discussions
New post →No discussions yet.