AI agent skill · by @Arize-ai

Phoenix Pr Screenshot

Screenshot a running Phoenix feature and attach images to a GitHub PR. Builds the frontend, starts Phoenix with env vars, captures browser screenshots, and attaches them to the PR body with `gh pr edit --attach`.

Archived 2026-10-23 · Source updated 2026-10-12. Check the original repository for the latest version.

SKILL.md instructions

# Phoenix PR Screenshot

Capture screenshots of the Phoenix UI to visually document a feature in a pull request. This skill handles the end-to-end workflow: build, launch, screenshot, upload, and attach to PR.

## Prerequisites

- Browser tooling capable of navigating the local Phoenix UI and saving screenshots
- `gh` CLI 2.102 or later (for `--attach`), authenticated with the Arize-ai/phoenix repo
- `pnpm` and `uv` available for building and running Phoenix

## Workflow

### Step 1: Build the frontend

The Phoenix backend serves the built frontend from `src/phoenix/server/static/`. Build it from the `js/app/` directory:

```bash
cd <repo-root>/app
pnpm install   # only if node_modules is missing
pnpm run build
```

This compiles the React app and copies static assets into the Python server's static directory. Without this step, page routes like `/playground` return 404.

### Step 2: Start Phoenix

Start the Phoenix backend with any env vars the feature requires. Always use a fresh working directory to avoid DB migration conflicts in worktrees:

```bash
PHOENIX_PORT=6007 PHOENIX_WORKING_DIR=/tmp/phoenix-screenshot-demo <OTHER_ENV_VARS> uv run phoenix serve &
```

Key points:
- Use `PHOENIX_PORT` (not `--port`) to set the port — the CLI doesn't accept a port flag
- Use a temp `PHOENIX_WORKING_DIR` so you don't collide with an existing DB that may have newer migrations
- Wait for the server to be ready: `sleep 10 && curl -s -o /dev/null -w "%{http_code}" http://localhost:6007/playground` should return 200
- Check `/tmp/phoenix-*.log` if it fails — common issues are migration errors (use a fresh working dir) or port conflicts

### Step 3: Capture screenshots

Use the available browser tooling to open `http://localhost:6007/playground` (or the relevant feature page), interact with the UI to show the feature, and save screenshots locally.

- Wait for the target UI elements to be visible and ready before interacting or capturing screenshots.
- Inspect the page again after navigation or DOM changes before choosing the next element to interact with.
- Take multiple screenshots when useful (before/after, dropdown open, etc.).
- For transient or animated states, take a burst of screenshots a fraction of a second apart while you trigger the state, keep the frames that show it, and assemble them into a GIF.
- View the saved screenshots to verify they captured what you intended, and use their local paths in the upload step.

### Step 4: Attach to the PR

`gh pr edit --attach` uploads an image or video to GitHub and rewrites any `./file` reference in the
body to the hosted URL. Name files `<PR_NUMBER>-<descriptive-name>.png` (e.g.
`11986-playground-loaded.png`). Read the current body first so nothing is lost:

```bash
gh pr view <PR_NUMBER> --json body -q .body > body.md
# Insert the media under the What sentence, with a bold caption stating what to notice:
#   **<what to notice>**
#
#   ![<alt text>](./<PR_NUMBER>-<name>.png)
gh pr edit <PR_NUMBER> --body-file body.md --attach './<PR_NUMBER>-<name>.png#<alt text>'
```

Run it from the directory holding the files, and pass `-R Arize-ai/phoenix` when that directory is
not a git checkout. `phoenix-pr-description` allows one media item, so for several states prefer a
single GIF or video over a series of stills.

### Step 5: Cleanup

```bash
# Kill the Phoenix server
kill <PID>
```

Close the browser session created for the screenshots.

## Removing screenshots

Read the body into a file as in Step 4, remove the caption and image reference, then apply it with `gh pr edit <PR_NUMBER> --body-file body.md`.