← Back to Skills Marketplace
newtontech

PR-Proof-of-Work

by Haoming Yan · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
92
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install pr-proof-of-work
Description
TDD-driven E2E workflow with real Playwright browser screenshots as PR proof. Use when: (1) fix bugs or implement features with test-first approach, (2) crea...
Usage Guidance
This skill is mostly what it says: it helps capture Playwright BEFORE/AFTER screenshots and post them on a PR. Before using it, check these things: (1) the SKILL.md and scripts assume you have git and the GitHub CLI (gh) installed and authenticated with push/comment permissions — the registry metadata did not declare these requirements; (2) the shell script will create a branch, commit screenshots into your repo, and run git push --force on a branch named e2e-screenshots-<PR_NUM> — be sure you are comfortable with that (it can overwrite remote branch history); (3) export E2E_SCREENSHOT_DIR or accept the default path; (4) example commands in the doc sometimes use jq and shuf — ensure those binaries exist if you use the examples; (5) review the scripts (assets/screenshot-reporter.ts and scripts/pr-comment-screenshots.sh) and run them in a safe/test repository or on a fork first to confirm behavior. If you need the skill to avoid modifying your repo, do not run the pr-comment script — instead upload images manually or adapt the script to your policies. If anything here is unclear, ask the maintainer for explicit dependency and permission documentation.
Capability Assessment
Purpose & Capability
Name/description align with included files: a TypeScript screenshot helper and a shell script that uploads screenshots to a PR. The files and SKILL.md are coherent with a TDD E2E screenshot-to-PR workflow. However, the skill metadata declares no required binaries or env vars while the instructions and script clearly rely on git, the GitHub CLI (gh), and optionally jq/shuf in examples — this mismatch is a notable omission.
Instruction Scope
SKILL.md stays on-task: it instructs creating tests, capturing BEFORE/AFTER screenshots, and making PRs. The provided shell script reads manifest.json, copies image files into the repo, creates a temporary branch, pushes images, and posts a PR comment via gh. All of these actions are within the described purpose and there is no hidden network exfiltration beyond pushing to the user's GitHub repo / using gh.
Install Mechanism
No install spec — instruction-only with two code files. No external downloads or archive extraction are present. This is low risk from an install-execution standpoint.
Credentials
The skill omits declaring required environment/binary dependencies: it uses E2E_SCREENSHOT_DIR (SKILL.md tells you to export it) and requires git and gh CLI with push permissions (and in one example uses jq/shuf). The script implicitly depends on an authenticated gh and write access to the repository remote. Those credentials/permissions are not listed in requires.env/primaryEnv, which is an important mismatch — the skill will act with your git/GitHub permissions if run.
Persistence & Privilege
The shell script creates a branch in the repository, copies screenshots into the repo, commits them and does a git push --force to a branch named e2e-screenshots-<PR>. It attempts to restore the user's current branch but performs remote force-pushes and modifies repo state. This behavior is coherent with the 'push screenshots and comment' purpose but is privileged (it mutates the user's repo and remote).
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install pr-proof-of-work
  3. After installation, invoke the skill by name or use /pr-proof-of-work
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release introducing a TDD-driven E2E PR workflow powered by Playwright screenshots as PR proof. - Define a strict "One issue → One PR" process with visual before/after proof for every code change. - Mandate studying passing tests and prohibit direct DOM modification via page.setContent(). - Require Playwright E2E tests using real browser, capturing BEFORE and AFTER screenshots. - Document setup, worktree isolation, screenshot management, and strict PR requirements. - Outline fallback strategies for backend-only changes by connecting code to observable UI outcomes. - Enforce a hard gate: both BEFORE/AFTER screenshots and passing tests required before opening a PR.
Metadata
Slug pr-proof-of-work
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is PR-Proof-of-Work?

TDD-driven E2E workflow with real Playwright browser screenshots as PR proof. Use when: (1) fix bugs or implement features with test-first approach, (2) crea... It is an AI Agent Skill for Claude Code / OpenClaw, with 92 downloads so far.

How do I install PR-Proof-of-Work?

Run "/install pr-proof-of-work" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is PR-Proof-of-Work free?

Yes, PR-Proof-of-Work is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does PR-Proof-of-Work support?

PR-Proof-of-Work is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created PR-Proof-of-Work?

It is built and maintained by Haoming Yan (@newtontech); the current version is v1.0.0.

💬 Comments