build-setup
Set up build environment and test fixtures for COBOL-to-Java migrations. Creates compilation infrastructure for both languages…
Set up the OpenHands automated PR review workflow in a GitHub repository. Creates the workflow file and asks the user for configuration preferences. Use when a user wants AI code review on their pull requests.
$ npx -y skills add openhands/extensions --skill setup-pr-review --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/setup-pr-reviewContext preview
The summary Claude sees to decide when to auto-load this skill.
Set up the OpenHands automated PR review workflow in a GitHub repository. Creates the workflow file and asks the user for configuration preferences. Use when a user wants AI code review on their pull requests.
name: setup-pr-review description: Set up the OpenHands automated PR review workflow in a GitHub repository. Creates the workflow file and asks the user for configuration preferences. Use when a user wants AI code review on their pull requests. triggers: - setup-pr-review - set up pr review - add code review workflow - openhands pr review
Add the PR review workflow to a GitHub repository so an OpenHands agent can review pull requests and post inline comments.
**Docs**: https://docs.all-hands.dev/sdk/guides/github-workflows/pr-review
Create `.github/workflows/pr-review.yml` in the target repo. Fetch the latest example from https://docs.all-hands.dev/sdk/guides/github-workflows/pr-review and use it as the starting template. The workflow calls the `OpenHands/extensions/plugins/pr-review` composite action directly.
Ask the user whether they are using the **OpenHands app** (app.all-hands.dev) or their **own LLM provider** (e.g. Anthropic, OpenAI directly).
OpenHands app users already have access to an LLM API key through the OpenHands litellm proxy. Tell them:
> Go to https://app.all-hands.dev → Account → API Keys → OpenHands LLM Key, and copy your key. > Then add it as a GitHub repository secret: > Settings → Secrets and variables → Actions → New repository secret. > Name it `LLM_API_KEY`.
Set these inputs in the workflow `with:` block:
If the user has their own API key (e.g. from Anthropic or OpenAI), tell them to add it as a repository secret named `LLM_API_KEY` using the same path above. Leave `llm-base-url` unset and set `llm-model` to the provider-prefixed model name (e.g. `anthropic/claude-sonnet-4-5-20250929`).
**You cannot create secrets — the user must do it manually.** Do not ask for the key value. Just tell them where to put it.
Present these options and apply any requested changes to the workflow file:
**Review style** (default: `roasted`)
simplicity.
**When to trigger** (default: on-demand only)
`on.pull_request.types` and matching conditions to the `if:` block.
After applying these, ask the user if they want to explore additional options (model selection, evidence requirements, custom review skills, observability). If yes, walk them through it — use the docs as a reference: https://docs.all-hands.dev/sdk/guides/github-workflows/pr-review If not, you're done.
This repository is the public extensions registry for OpenHands. It contains reusable, shareable skills, plugins, and Agent Canvas Apps that customize agent behavior and interfaces. Skills overview docs: SDK skill guide:
Set up build environment and test fixtures for COBOL-to-Java migrations. Creates compilation infrastructure for both languages…
End-to-end COBOL to Java migration workflow. Handles build setup, mainframe dependency removal, and code migration with test…
Create a transformation guide for replacing mainframe-specific COBOL constructs with standard COBOL equivalents. Use when…
Apply mainframe dependency transformations to COBOL code using a pre-generated transformation guide. Converts CICS/VSAM…
Migrate COBOL code to idiomatic Java, preserving business logic while following Java best practices. Use for COBOL modernization…
A test skill that responds to the magic word "alakazam" with a specific phrase