Prox OS Internal Docs
DevelopmentAI Control PlaneRuns

Agent Run Logs

Every meaningful agent execution should leave a run log. Run logs make it

Purpose

Every meaningful agent execution should leave a run log. Run logs make it possible to reconstruct what prompt, scope, branch, files, commands, risks, and follow-up came from each worker.

Naming

YYYY-MM-DD-short-kebab-title.md

Template

# YYYY-MM-DD Agent Run

Run ID:
Date:
Agent:
Source Prompt ID:
Branch:
Task Graph:
Result:

## Scope

## Changed Files

## Validation Commands Run

## Result Notes

## Risks

## Follow-up

Result Values

completed
partial
failed
blocked

On this page