Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
19 changes: 13 additions & 6 deletions README.md
Original file line number Diff line number Diff line change
@@ -1,4 +1,7 @@
# Humane Gate
# Humane Gate (demo repo)

> **This repo is a demo, not the thing to install.** It exists so you can read real example pull requests and their verdicts before adopting anything. To use Humane Gate on your own repositories, install [**humane-gate-action**](https://github.com/buildinghumanetech/humane-gate-action). It is one secret and one workflow file.


A HumaneBench check that runs on every pull request, scores the diff against
**HumaneBench rubric v4**, and **blocks nothing**.
Expand Down Expand Up @@ -47,9 +50,11 @@ A finding has to survive three filters before anyone sees it:

Anything else is dropped by the runner, not by the model.

## Setup
## Run this demo yourself

For maintainers and the curious. Adopters do not need any of this; use [humane-gate-action](https://github.com/buildinghumanetech/humane-gate-action).

1. New public repo, push this.
1. Create a new public repo and push this one to it.
2. **`./scripts/sync_rubric.sh --check`**: the checked-in rubric is already
byte-identical to upstream and pinned in `rubrics/VERSION`. This confirms it
has not drifted. Without `--check` it pulls the new bytes and re-pins.
Expand Down Expand Up @@ -206,15 +211,17 @@ turn those on. Credit to the Sparkle implementation for the two-checkout design.
what would actually fix it.
- Set `HUMANEBENCH_MODEL` to whichever model you have access to.
- Secrets are not available to `pull_request` runs from forks. Fine for a demo
repo you own; a real deployment uses `pull_request_target` with a pinned
checkout, or a GitHub App.
repo you own. Do not use `pull_request_target` to work around that: it runs
with secrets against fork code. Real deployments should use
[humane-gate-action](https://github.com/buildinghumanetech/humane-gate-action),
which does not support forks in v1 for exactly this reason.
- The check run is hardcoded to `conclusion: neutral`. Making it gate is a
one-line change, and that line is the thirty-day decision.

## Licence

Apache License 2.0. See [`LICENSE`](LICENSE). Copyright 2026 Building Humane
Technology.
Tech.

`rubrics/rubric_v4.md` is vendored from
[buildinghumanetech/humanebench](https://github.com/buildinghumanetech/humanebench)
Expand Down
Loading