How do I install the Test Plan skill?

Run npx skillmds@latest add hoophq/test-plan in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

What does the Test Plan skill do?

Generate the "How to test" section for the current branch's changes - exact commands, flows to exercise, expected results. Use before opening/updating a PR or when asked how to validate a change. It is listed under Coding & Dev Tools on SkillMD.

Is Test Plan safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Test Plan?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Test Plan free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Test Plan?

hoophq (@hoophq) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Test Plan by hoophq — Agent Skill

Name: Test Plan
Author: hoophq

File contents

Generate a test plan for the current changes

Goal: a teammate (or a QA agent) must be able to validate this change without asking anything. This is the team's standard for every PR.

Steps

Diff the branch against main: git diff main...HEAD --stat then read the changed files' diffs.
Identify what is observable from outside: which API response, CLI behavior (hoop connect / hoop exec / proxy flow), gateway/agent interaction, or webapp_v2 screen changed.
Write a ## How to test section in markdown:
- Setup: exact commands — see DEV.md (make run-dev, Postgres via make run-dev-postgres, agent, webapp_v2: npm run dev). Only what this change needs.
- Steps: numbered, copy-pasteable commands or precise UI actions.
- Expected: the observable result per step — literal output where possible (API status + body fields, CLI output, what the screen shows).
- Regression check: 1-2 quick checks that the closest existing behavior still works.
Cover the negative path too (permission denied, invalid input, guardrail still blocking after a fix).
If a step cannot be validated locally (needs enterprise libhoop, IDP, cloud resources), say so explicitly and state what environment is needed — never leave silent gaps.

Output the section ready to paste into the PR body (or update the open PR with gh pr edit --body-file if asked).

Frequently asked questions

Test Plan