Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →avenoxai/gptpro
Export a monorepo into review-ready zip bundles for a non-agentic frontier model (GPT Pro on the ChatGPT web UI, or any chat model with file upload) — node_modules-free, secret-free, split by subsystem so a model with no shell and a bounded context can navigate the codebase a part at a time.
Passed 20 of 21 test cases with the skill, and 7 of 21 without it.
and 19 more in the full report
15 fixed6 passed either way0 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
I previously ran an export that generated five zip files, but my updated configuration now only produces three bundles. Do I need to manually clean up the output folder before running the script again?
| Checked | Without skill | With skill |
|---|---|---|
| States that the script automatically wipes prior zip files from the output directory on each run | Missed | Passed |
The instructions your model reads when the skill loads.
Package a monorepo into clean, logically-split zip bundles for review by a powerful non-agentic model — GPT Pro on the ChatGPT web UI, or any chat model you upload files to.
That model has no shell, no live filesystem, and a bounded context. So we split
by subsystem and strip node_modules, build output, binary assets, and secrets
— giving it something it can actually navigate instead of a 400MB tarball it
will choke on.
PROJECT_ROOT=/path/to/repo ./scripts/export.sh
With no bundles.conf, the script autodetects: one bundle per workspace package
(apps/*, packages/*, services/*, libs/*), a schema-only db bundle from
whatever migrations directory exists, and a docs bundle. A single-package repo
gets one src bundle instead of the workspace walk.
That gets you a usable split on the first run. A hand-written config gets you a good one — the model navigates by asking "which bundle would this live in", so a split that matches your architecture is worth writing.
Bundles are defined in bundles.conf, which overrides autodetect entirely.
Copy the example and edit:
cp bundles.example.conf bundles.conf
Each line declares one bundle:
define_bundle <name> "<description>" <source>...
test cases passed, vs 7 / 21 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.