@@ -0,0 +1,44 @@
+name: PR Limit
+
+# Open-PR limit for contributors (see CONTRIBUTING.md, "Pull Request Limits and Review
+# Bandwidth"). The logic, limits and wording live in the shared reusable workflow in
+# floci-io/.github; this file only wires the events to it.
+#
+# pull_request_target gives the token write access on fork PRs. Neither this workflow
+# nor the reusable one checks out or runs PR code; they only call the GitHub API.
+
+on:
+  pull_request_target:
+    # closed and converted_to_draft only remove stale over-pr-limit labels.
+    types: [opened, reopened, ready_for_review, closed, converted_to_draft]
+  # Manual hook: evaluate any open PR. Dry run by default, so it only logs the decision.
+  workflow_dispatch:
+    inputs:
+      pr_number:
+        description: 'PR number to evaluate'
+        required: true
+      dry_run:
+        description: 'Only log what would happen'
+        type: boolean
+        default: true
+
+permissions:
+  issues: write
+  pull-requests: write
+
+concurrency:
+  # github.ref is the base branch under pull_request_target, so key on the PR number;
+  # otherwise every PR run would cancel every other. Manual runs get their own group:
+  # GitHub keeps only one pending run per group, so sharing it would let a manual dry
+  # run replace a live evaluation waiting in the queue. Live runs for the same PR queue
+  # rather than cancel; a newer live event may replace an older pending one, which is
+  # fine because it evaluates the PR's current state.
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || inputs.pr_number }}-${{ github.event_name == 'workflow_dispatch' && 'manual' || 'live' }}
+  cancel-in-progress: false
+
+jobs:
+  pr-limit:
+    uses: floci-io/.github/.github/workflows/pr-limit.yml@main
+    with:
+      pr-number: ${{ format('{0}', github.event.pull_request.number || inputs.pr_number) }}
+      dry-run: ${{ github.event_name == 'workflow_dispatch' && inputs.dry_run }}
@@ -0,0 +1,84 @@
+# Contributing to testcontainers-floci (Python)
+
+Thanks for contributing. Issues, reviews, docs, and pull requests are all welcome, and you
+don't need permission to start. Before you dig in, a quick read of the project
+[GOVERNANCE.md](https://github.com/floci-io/.github/blob/main/GOVERNANCE.md) explains how
+decisions are made and how contributors grow into Maintainers over time.
+
+## Developer Certificate of Origin (DCO) sign-off
+
+Every commit must be **signed off**, certifying the
+[Developer Certificate of Origin](https://developercertificate.org/), a lightweight statement
+that you wrote the contribution or otherwise have the right to submit it under the project's
+license. It is **required for a pull request to be merged**.
+
+Sign off by adding the `-s` flag when you commit:
+
+```bash
+git commit -s -m "feat(config): add Kinesis service configuration"
+```
+
+This appends a `Signed-off-by: Your Name <your@email>` trailer using your configured git
+identity. If you forget, you can amend the most recent commit with `git commit --amend -s`, or
+sign off a range during an interactive rebase.
+
+Why Floci uses the DCO and not a CLA is explained in the
+[organization contributing guide](https://github.com/floci-io/.github/blob/main/CONTRIBUTING.md#why-the-dco-and-not-a-cla).
+
+## Commit messages
+
+- Use [Conventional Commits](https://www.conventionalcommits.org/), e.g.
+  `fix(container): expose the RDS port range`.
+- Releases are cut by semantic-release from these messages: `feat` and `fix` commits decide the
+  next version, so pick the type that matches the change.
+- **No AI attribution.** Do not add "Generated by", "Co-Authored-By: …-bot", or similar
+  trailers to commit messages.
+
+## Running tests locally
+
+Set up a development environment with [uv](https://docs.astral.sh/uv/) (or
+`pip install -e ".[dev]"`):
+
+```bash
+uv sync --extra dev
+```
+
+Lint, format check and type check, as CI runs them:
+
+```bash
+ruff check .
+ruff format --check .
+mypy floci
+```
+
+Unit tests need no Docker:
+
+```bash
+pytest -m "not integration" -v
+```
+
+Integration tests start a real Floci container, so they need Docker:
+
+```bash
+docker pull floci/floci:latest
+pytest -m integration -v
+```
+
+## Pull Request Limits and Review Bandwidth
+
+To make sure every contribution gets a thorough, high-quality review in a reasonable time, we ask contributors to keep **no more than 2 open, non-draft pull requests** at any time in this repository.
+
+- **Why this policy exists:** maintainer review time is limited. Capping concurrent open PRs prevents review backlogs, reduces context switching, and keeps PR cycle times short for everyone.
+- **Dependent work:** if your work depends on a PR that has not been merged yet, build on that branch or note the dependency in the discussion instead of opening separate, uncoordinated PRs.
+- **Draft PRs:** drafts do not count against the limit. Mark a draft as ready for review only when you have review capacity available.
+- **How it is applied:** a bot labels your 3rd and later open pull requests `over-pr-limit` with a reminder. Starting **2026-10-08**, from your 5th open pull request onward, new ones are closed automatically. Your branch and commits are kept, and you can reopen the PR once one of your other PRs is merged or closed. Maintainers and dependency bots are not counted.
+
+Once your current pull requests are reviewed, merged, or closed, you are welcome to open new ones!
+
+## Questions and discussion
+
+- 💬 **[Slack](https://join.slack.com/t/floci/shared_invite/zt-3tjn02s3q-A00kEjJ1cZxsg_imTfy6Cw)**: quick questions and community chat.
+- 🗣️ **[GitHub Discussions](https://github.com/orgs/floci-io/discussions)**: feature ideas, design tradeoffs, and proposals.
+
+By contributing, you agree to abide by the project
+[Code of Conduct](https://github.com/floci-io/.github/blob/main/CODE_OF_CONDUCT.md).