From browserctl
Capture a browserctl product feedback entry as a local markdown file. Use when the user reports a bug, friction, missing feature, or doc gap in browserctl — phrases like "this is broken", "this should work differently", "the docs don't mention", "I wish browserctl could", or an explicit "/feedback".
How this skill is triggered — by the user, by Claude, or both
Slash command
/browserctl:feedbackThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Save the user's feedback about browserctl as a structured markdown file in `.claude/feedback/` so it can be filed as a GitHub issue later without re-asking the user to repeat themselves.
Save the user's feedback about browserctl as a structured markdown file in .claude/feedback/ so it can be filed as a GitHub issue later without re-asking the user to repeat themselves.
Trigger on any of:
/feedback (with or without arguments).browserctl, browserd, or a workflow file — even mid-task.If the user is mid-task and you spot a feedback-worthy moment, capture it without derailing the task: write the file, mention the path, then continue.
The user-supplied content (if any) appears as $ARGUMENTS. If empty, ask one focused question:
"What's the feedback? Describe the issue, friction, or idea — include what you were doing when you noticed it."
Tag every entry with one or more types:
| Type | When to use |
|---|---|
bug | Something behaved incorrectly or errored unexpectedly |
ux | Something worked but felt awkward, slow, or confusing |
feature | Something is missing that would make the tool more useful |
docs | Documentation was wrong, missing, or misleading |
A single entry may cover more than one type — list all that apply.
Pull as much as possible from the current conversation without asking the user to repeat themselves:
If critical information is missing and can't be inferred, ask one focused question — not a list.
.claude/feedback/YYYY-MM-DD-<slug>.md<slug>: 3–5 word kebab-case summary (e.g. click-js-button-no-fire).File template:
# <short title>
**Date:** YYYY-MM-DD
**Type:** bug | ux | feature | docs (list all that apply)
**Area:** <the command or concept this relates to, e.g. `click`, `snapshot`, `workflow DSL`, `daemon`, `driver`>
## Summary
One or two sentences clear enough to file as a GitHub issue title + body.
## Context
What the user was doing when this came up. Include the exact command(s) if relevant.
## Expected behaviour
What should have happened.
## Actual behaviour
What happened instead. Include error output or screenshot paths if captured.
## Workaround
How to work around it today (if one exists).
## Suggested fix or improvement
Concrete suggestion — a wording change, new flag, changed default, new command, etc. Leave blank if unknown.
## Notes
Anything else useful for the maintainer.
Omit sections that don't apply (e.g. "Actual behaviour" for a pure feature request).
Tell the user:
Do not open a browser, create a GitHub issue, or take any external action unless the user explicitly asks.
Guides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.
npx claudepluginhub patrick204nqh/browserctl --plugin browserctl