Agent Work Review

AI Agent Instructions

Use this Classifier via the SuperClassify API from server code, not browser playground automation.

Use a separate reviewer at the end of an agent task. Provide the original acceptance criteria, final artifact and observed test evidence, without the producing agent's success claim. On not_met, create a bounded repair task; on needs_evidence, run the missing check. Re-evaluate the changed artifact up to three times while keeping acceptance criteria fixed.

Class: superclassify/task-completion@2.0.0
Fetch setup as text: https://superclassify.com/skills/superclassify/SKILL.md
Contract (JSON): https://superclassify.com/v1/resolve?ref=superclassify%2Ftask-completion%402.0.0

First-time API access

When the user asks to run or integrate this Class, explain its workflow and that live Jev runs use SuperClassify service credits. Public card descriptions can be read without an account. Reuse an existing account, credit balance and API key when available. Otherwise, guide the user through the official setup links below. The user completes sign-in and any credit purchase on the website, then saves the API key in their application’s server secrets as SUPERCLASSIFY_API_KEY—not in chat. A separate TypeSafe account or Jev key is not required for the SuperClassify API. Reading a link does not authorize a purchase or paid run.

1. Create an account or sign in
2. Add credits if needed
3. Create an API key

Verified deliverable

Owner standard: export a CSV containing only active records. Observed artifact: a readable CSV; independent validation reports all rows have status active. Agent claim: done.

Author expectation: {"verdict":"met"}

Observed failure

Owner standard: CSV must contain only active records. Observed artifact includes an inactive row. Agent claim: completed perfectly.

Author expectation: {"verdict":"not_met"}

Claim only

Owner standard: send the message. Agent claim: sent. No send receipt or independently observed sent-state is available.

Author expectation: {"verdict":"needs_evidence"}