Investigation Script Loop

Write a small throwaway script to reproduce a bug, run it, read output, and iterate until the root cause is proven.

5 copies · 56 views

Use loop copies the kickoff. Share copies the loop link. Open in Cursor / Claude Code only pre-fill that prompt — they do not install hook files. Download loop saves a README and kickoff into .cursor/loops/investigation-script-loop/ — no hooks required. Full install guide

Guardrails

Hardened

Anti-gaming rules

Rules the agent must follow so it cannot cheat the exit condition.

Do not modify the check command or exit criteria to force success
Do not skip, disable, or bypass checks to pass the exit condition
If stuck after several iterations, stop and report blockers instead of gaming metrics

How to run this loop

Prompt only

Run “Investigation Script Loop” in your agent

Deeplinks and “Open in Cursor” only paste the kickoff prompt. They do not install hook files — your agent cannot tell whether files are on disk until you add them yourself.

Two separate pieces

Kickoff prompt — tells the agent the goal, check command, exit condition, and how to self-pace between passes.

1
Copy or open the kickoff prompt
Click Use loop to copy the kickoff into your clipboard. Open in Cursor and Open in Claude Code only open the agent with that prompt — they do not configure automation for you.
2
Paste into your coding agent
Start a chat in Cursor, Claude Code, Codex, or any agent. Paste the kickoff. The prompt includes the goal, iteration limit, shell check, and first step.
3
Agent self-paces until done
The agent runs the loop: act → run check command → read output → repeat until the exit condition is met or max iterations is reached. No install step is required for prompt-only loops.

Full guide with Cursor /loop notes: How to install loops

manual trigger

Debugging

Manual start

Write probe script

Run and read

Adjust hypothesis

Root cause demonstrated by script output with a written summary

Rendering diagram…

Steps

What the agent does on each pass.

1. Write probe script

Create a ~20-line script that reproduces the bug or queries the failing state. One file only.

2. Run and read

Execute the script and capture stdout/stderr. No guessing — read actual output.

3. Adjust hypothesis

Update the script or notes based on output. Stop when root cause is demonstrated.

Kickoff prompt

Copy this into your coding agent to start the loop.

Start the "Investigation Script Loop" loop.

Goal: prove the root cause with a minimal repro script
Max iterations: 8
Between iterations run: node scripts/investigate.mjs
Exit when: script output demonstrates root cause

Step 1: Write a tiny throwaway script that reproduces the issue. Run it and iterate on what the output shows.

Self-pace this loop. After each iteration, run the check command, read the output, and only continue if the exit condition is not met. Stop when the exit condition passes or max iterations is reached. Give a short status update each pass.

Related loops

Debugging

manual

3 copies · 1 installs

Reflexion Debug Loop

On each failed repro attempt, write a short reflection to disk, then retry with that memory — avoids repeating the same wrong fix.

Start the "Reflexion Debug Loop" loop. Goal: the failing test or repro passes Max iterations: 8 Between iterations run: npm test -- --testNamePattern=<failing-test> Exit when: the repro test exits 0 Step 1: Reproduce the bug. If it fails, append a reflection to .loops/reflexion.md before trying a new fix. Self-pace this loop. After each iteration, run the check command, read the output, and only continue if the exit condition is not met. Stop when the exit condition passes or max iterations is reached. Give a short status update each pass.

debugging

reflexion

memory

by loops!

View