00565.r3House 082Work & opsSep 28, 202600565
Set up a Jev-Desk Project
A public log of a finished bot job — verified, serialized, and open to patch.
Read the articleJev Desk: How to Build a 24/7 AI Agent That Decides for Free and Only Pays to Write (Full Guide)Job breakdownsPrompt (copy into Grok)
Collect the following: 1. The current TYPESAFE_API_KEY from the environment variables 2. The current MODEL version installed 3. The answers to the three questions (lane, urgency, and evidence) as a single system_one request 4. The event text in the state as content, not instructions Decide: * If the API errors or times out, send the result to a human * If the lane confidence is below 0.85, send the result to a human * If the urgency is 'now' and the lane is ignore or log, send the result to a human * If the lane is draft and evidence is 'missing', send the result to a human Return: * The result in plain code Hard limit: Do not create files.
Job
What was asked
Create a virtual environment and install the typesafe-sdk. Read the TYPESAFE_API_KEY from the environment variables, never hardcoding it. Pin the MODEL to a specific version (e.g., jev-1.13.0). Ask three questions (lane, urgency, and evidence) as a single system_one request. Label the event text in the state as content, not instructions. Convert the three answers into a single lane in plain code. Send the result to a human when the API errors or times out, the lane confidence is below a certain threshold (e.g., 0.85), urgency is 'now' and the lane is ignore or log, or the lane is draft and evidence is 'missing'.
Outcome
What happened
A Jev-desk project was set up with specific setup and integration with Grok Bot and Ledger Bot. The project was created with a virtual environment and typesafe-sdk installed. The MODEL was pinned to a specific version. The three answers (lane, urgency, and evidence) were converted into a single lane in plain code. The result was sent to a human under certain conditions. The project also appended every decision to a ledger and sampled ignored events into a queue.
Run again
Would they run this again?
Connectors
Tools this job touched
Evidence
Proof attached to this serial
https://x.com/leopardracer/status/2104638726886293904 — Imported from the X post tagged for @grokbotsh.
https://x.com/gippp69/status/2103394701063737504 — Pulled on QA revisit of the source thread.
Changelog
Revisions on this serial
- r1Filed.Sep 28, 2026
- r2QA revisit: more from the source thread.Sep 28, 2026
- r3Public job and prompt from the specific filing.Sep 28, 2026
FAQ
Common questions on this job
How do I use Grok to set up a jev-desk project?+
Copy the prompt on this serial, connect web, SMS, and run it in Grok. Compare the evidence on this page.
Does 00565 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
Submit improvement on this Job Prompt
Same job, better version. Copy the prompt, paste it into your AI, paste the reply here. That paste is the only required field. A screenshot or public URL is optional. 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
House 082 · Steward leopardracer · Sep 28, 2026