Let an LLM finish browser tasks through a harness
Built Browser Harness so an LLM can complete arbitrary browser tasks with fewer guardrail dead-ends.
Source
gregpr07 on Hacker News · 2026-04-24 · Project link
Written in our own words. Credit to the original author. Link checked 2026-10-15.
What they did
Built Browser Harness so an LLM can complete arbitrary browser tasks with fewer guardrail dead-ends.
Result (as claimed)
Show HN browser-agent harness.
Tools they used
Cost: OpenAI Codex: free plan, paid from $8/mo. Checked 2026-10-14.
Set it up
Map the task
Write the trigger, the inputs, the decisions and the output. Note which steps need a human yes.
Connect the tools
Set up Claude, ChatGPT, OpenAI Codex with the narrowest permissions that work. Use a test account or sandbox first.
Build the workflow
Create the steps in order. Add an approval step before anything sends, pays or deletes.
Run it on real but low-stakes data
Run a few times, read every output, and fix prompts or rules that misfire.
Schedule and monitor
Turn on the schedule or trigger, and add a daily summary or alert so you notice failures.
Copy the full setup prompt
Paste into your agent and it will walk you through the setup, asking before anything that sends, posts or spends.
You are helping me set up this AI use case end to end in Claude Code. Goal: Let an LLM finish browser tasks through a harness What the original author did: Built Browser Harness so an LLM can complete arbitrary browser tasks with fewer guardrail dead-ends. Reported result (as claimed by the author): Show HN browser-agent harness. Source: https://news.ycombinator.com/item?id=47890841 Tools they used: Claude, ChatGPT, OpenAI Codex Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
You are helping me set up this AI use case end to end in Codex. Goal: Let an LLM finish browser tasks through a harness What the original author did: Built Browser Harness so an LLM can complete arbitrary browser tasks with fewer guardrail dead-ends. Reported result (as claimed by the author): Show HN browser-agent harness. Source: https://news.ycombinator.com/item?id=47890841 Tools they used: Claude, ChatGPT, OpenAI Codex Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
You are helping me set up this AI use case end to end in Grok Bot. Goal: Let an LLM finish browser tasks through a harness What the original author did: Built Browser Harness so an LLM can complete arbitrary browser tasks with fewer guardrail dead-ends. Reported result (as claimed by the author): Show HN browser-agent harness. Source: https://news.ycombinator.com/item?id=47890841 Tools they used: Claude, ChatGPT, OpenAI Codex Category: Automation & business Please: 1. Ask me 3 short questions about my situation (accounts I have, budget, what I want the output to look like). 2. Propose a step-by-step plan using these tools or cheaper alternatives, with the time and cost of each step. Use official pricing pages; say "check pricing" if unsure. 3. Then do the setup with me one step at a time, waiting for my OK before anything that sends messages, posts publicly, spends money or deletes data. 4. Finish with a checklist to run it again next time.
The Daily 5
Get five new use cases tomorrow morning
Free. Real sources. Unsubscribe any time.
Related use cases
Find which LLM calls a cheaper model could have handled
Built Frugon, a local MIT tool that finds which LLM calls a cheaper model could have handled.
Build a ChatGPT and Claude plugin that applies to jobs
Built a ChatGPT and Claude plugin that helps apply to jobs from inside the chat interface.
Automate podcast guest signups
Connect Calendly to Make so guest bookings flow into the rest of the podcast workflow automatically.
Hand repeatable business routines to an agent
A founder running $20M+ businesses rates Grok Bot top tier because it reliably runs repeatable routines.