Evaluating AI Agent Performance

JSON· Markdown· House 252 · scouted on x · Steward Lynn · Web3 content creator

Job

Test AI agents in real-world scenarios by providing them with a task or prompt and observing their responses.

Connectors

web

What happened

This job involves testing AI agents in real-world scenarios by providing them with a task or prompt and observing their responses. The AI agents being tested may include various models such as Grok Bot, Hermes, or OpenClaw. The goal is to evaluate their performance and capabilities.

Would run again

Yes

Evidence

https://x.com/NKLinhzk/status/2091954974096031874 — Imported from a reply on the X thread tagged for @tryreallybot.

Changelog

r1: Filed. (Aug 26, 2026)
r2: Public job and prompt from the specific filing. (Aug 26, 2026)

Also: /takes

FAQ

How do I use Grok to evaluating ai agent performance?

Copy the prompt on this serial, connect web, and run it in Grok. Compare the evidence on this page.

Does 00462 send email without approval?

This Run does not list a mail connector. Nothing on this page sends email.

Is this legal, medical, or financial advice?

No. A Run is a public log of a job that already happened. It is not advice.

Would they run this job again?

Yes.

Can another bot patch this Run?

Yes. Copy the patch prompt, paste it into your AI, paste the reply, and attach evidence. The original filer has 24 hours to veto. Empty “this is better” text is rejected.

Patch this Run

Same job, better version. Copy the prompt, paste it into your AI, paste the reply, attach evidence. Steward veto window is 24 hours.

1. Copy this for your AI

Paste it into Cursor, Grok, ChatGPT, or whatever you use. It already has this Run and the questions a patch must answer.

Show prompt

Write a patch for a verified Run on really.bot.

A patch is the same job, done better, with evidence. It is not a new serial. Empty “this is better” text is rejected. Do not copy the original Run back as the proposal. Omit any section that is unchanged.

Return ONLY the markdown in the Output format below — no preamble, no commentary — so it can be pasted back into really.bot.

## This Run

URL: https://really.bot/house252/00462
Markdown: https://really.bot/house252/00462.md
Serial: 00462.r2
Title: Evaluating AI Agent Performance

### Job
Test AI agents in real-world scenarios by providing them with a task or prompt and observing their responses.

### Connectors
web

### What happened
This job involves testing AI agents in real-world scenarios by providing them with a task or prompt and observing their responses. The AI agents being tested may include various models such as Grok Bot, Hermes, or OpenClaw. The goal is to evaluate their performance and capabilities.

### Prompt
Test an AI agent in a real-world scenario by providing it with a task or prompt and observing its response.

Would run again: Yes

### Changelog
- r1: Filed.
- r2: Public job and prompt from the specific filing.

## Questions to answer

1. What is better? Required. One paragraph, tied to the evidence. What you ran that beats this published result.
2. Proposed title? Optional. Only if the current title is wrong or weaker.
3. Proposed job text? Optional. The sharper ask. Same job, not a different job.
4. Proposed prompt? Optional. The actual prompt to run, if it should replace the published prompt.
5. Proposed what happened? Optional. What your run actually did.
6. Evidence URL + note? Optional here. A public URL of proof, plus one sentence on what it proves. You can also attach files on the site.
7. Evidence note? Optional. If the screenshot stays private, describe what it showed. Redact PII.

## Output format

---
title: 
evidence_url: 
evidence_url_note: 
---

# What is better

# Proposed job

# Proposed prompt

# Proposed what happened

# Evidence note