Skip to content
AI & Agents
Command

/feature-request

Draft a feature request issue for lean4-skills

From plugin
lean4-skills
36214 skills4 agents14 commands
Install
> /plugin marketplace add cameronfreer/lean4-skills

How it fires

How this command gets triggered: by you, by Claude, or both.

  • Fires itselfClaude auto-loads it when your prompt matches the work.
  • You can call itInvoke it directly when you want it.
  • Slash command/feature-request

Context preview

What this command does when you run it.

Draft a feature request issue for lean4-skills

Command definition

feature-request.md
name: feature-request
description: Draft a feature request issue for lean4-skills

Feature Request

Draft and submit a feature request as a GitHub issue on `cameronfreer/lean4-skills`.

Precondition

Invoke this command only if the user explicitly ran it or explicitly agreed to draft a feature request. If invoked without prior opt-in, ask:

> This looks like a plugin workflow gap. Want me to draft a feature request?

If the user explicitly ran `/lean4-contribute:feature-request`, treat that as opt-in. Do not proceed unless the user confirms. Do not gather structured questions or build an issue body until consent is given.

Gathering Context

Collect the following from the user and the current session. Ask for anything you cannot infer:

1. **Problem / pain point** — What is frustrating, slow, or missing? 2. **Current workaround** — How do you handle this today? (or "none") 3. **Desired behavior** — What should the plugin do instead? 4. **Why this belongs in the plugin** — Why is this a plugin concern rather than a user-side script or external tool? 5. **Acceptance criteria** — Concrete conditions that would mark this done 6. **Alternatives considered** — Other approaches you thought of and why they fall short 7. **Urgency / frequency** — How often do you hit this? Blocking, frequent annoyance, nice-to-have?

Drafting the Issue

Compose the issue using this template:

Title: [Feature] <short description>
Labels: enhancement

## Problem / Pain Point
<problem description>

## Current Workaround
<workaround or "none">

## Desired Behavior
<what the plugin should do>

## Why This Belongs in the Plugin
<justification>

## Acceptance Criteria
- [ ] <criterion 1>
- [ ] <criterion 2>
- ...

## Alternatives Considered
<other approaches and trade-offs>

## Urgency / Frequency
<how often, how blocking>

---
*Drafted via [lean4-contribute](https://github.com/cameronfreer/lean4-skills)*

Showing the Draft

Display the **complete** issue draft to the user — title, labels, and full body. Ask:

> Here is the feature request I will submit. Review it carefully. > > By submitting, you agree that this content may be edited, rearranged, or > incorporated into lean4-skills in any form under its > [MIT license](https://github.com/cameronfreer/lean4-skills/blob/main/LICENSE). > > **Submit this issue?** (yes / edit / cancel)

Do **not** proceed unless the user explicitly confirms.

Submitting

After confirmation, submit using the first available method:

1. **`gh` CLI** — Try: `gh issue create --repo cameronfreer/lean4-skills --title "<title>" --body "<body>" --label enhancement`. If the label fails (e.g. `enhancement` doesn't exist on the repo), retry without `--label` and note that the label was advisory only. If `gh` fails for auth or network reasons, **stop** — do not retry. Fall through to the browser or email path immediately. 2. **Browser fallback** — Provide a prefilled GitHub URL: `https://github.com/cameronfreer/lean4-skills/issues/new?title=<url-encoded-title>&body=<url-encoded-body>&labels=enhancement` (the `labels=` param is best-effort; GitHub ignores unknown labels silently) 3. **Email fallback** — Draft an email to `lean4skills@gmail.com` with subject `[Feature] <short description>` and the full issue body, for the user to send manually.

Report the result (issue URL, fallback URL, or email draft) and confirm completion.

Read more
Ships withlean4-skills

Lean 4 workflow pack for AI coding agents. Gives your agent a structured prove/review/golf loop, mathlib search, axiom checking, and safety guardrails.

Get the whole plugin