[EMSR-1912] Add Qualitative Feedback Setup skill

[fsaresh]

Summary

Adds a new agent skill, launchdarkly-flag-qualitative-feedback-setup, that walks users through wiring LaunchDarkly qualitative user feedback into a JS/TS codebase.

The skill is a step-by-step wizard, where it:

• verifies the project has a compatible client-side SDK (v3.0+),
• detects the framework and design system,
• gathers requirements (flag, prompt, widget style, placement),
• confirms the plan, then
• generates a sendFeedback utility and a feedback widget that emit the $ld:feedback event and match the project's existing patterns.

It also includes TypeScript templates (sendFeedback.ts, PopoverFeedback.tsx, InlineFeedback.tsx) and client/server SDK reference lists, and slots into the existing feature-flags skill family (let me know if you disagree and we can chat placement!).

Testing

• Not applicable
• Manual (describe below)
  • Ran the skill end-to-end against a Next.js / React demo app across multiple iterations, verifying generated code builds and feedback events reach LaunchDarkly.
    • Environment: Next.js 16 + React 19 app with Tailwind, launchdarkly-react-client-sdk v3.9, and @launchdarkly/observability + @launchdarkly/session-replay installed. Client-side SDK gate, design-system detection, and session-replay linkage were all exercised.
    • Scenarios: raw skill invocation, partial and full natural-language prompts, and a follow-up request adding a second feedback widget to a different flag in the same conversation. Covered smiley-face and thumbs-plus-text popover styles across separate flags/pages.
    • Verification each run: rebuilt the app (next build + container rebuild), confirmed the widget rendered at the chosen placement, the $ld:feedback event fired, and feedback surfaced on the flag's dashboard.
    • Repo checks: validate_skills.py (49 skills), generate_catalog.py --check (catalog current), and the unit test suite all pass.

Notes

This skill was authored in a separate repo, so this PR also packages it to match ai-tooling conventions.

• This does lose some of the [git history associated with the iterations](https://github.com/launchdarkly-labs/qualitative-feedback-setup-skill/commits/fsareshwala/productionize_qf_feedback_skill/), but I doubt those are super valuable here either way.
• Several of the skill's behaviors were shaped by the iterative test runs above, e.g. the explicit "step-by-step wizard" framing, the numbered Check 1/4–4/4 gates, the labeled confirmation selection, and the sentiment-vs-text widget-style question.

---

Note

Low Risk
Documentation and agent skill assets only; no runtime or production app code changes in this repo.

Overview
Adds launchdarkly-flag-qualitative-feedback-setup to the feature-flags skill family and registers it in README.md and skills.json.

The new package is a step-by-step agent wizard (SKILL.md) that gates on client-side LaunchDarkly JS SDK v3+, collects flag/prompt/widget/placement choices, uses MCP to verify or create the flag, requires explicit plan approval, then guides codegen and end-to-end verification of $ld:feedback events.

It ships reference templates (sendFeedback.ts, PopoverFeedback.tsx, InlineFeedback.tsx) plus client/server SDK allowlists, with marketplace.json declaring the LaunchDarkly MCP dependency.

^{Reviewed by Cursor Bugbot for commit 85d097d. Bugbot is set up for automated code reviews on this repo. Configure here.}

[erangeles]

@fsaresh can you please add a video or recording of the skill for context please?

[github-actions]

Skill eval results

Skill	Before	After	Δ
agentcontrol/configs-create	100/100 (4/4)	75/100 (3/4)	-25
agentcontrol/configs-update	80/100 (4/5)	80/100 (4/5)	no change
agentcontrol/configs-variations	80/100 (4/5)	80/100 (4/5)	no change
agentcontrol/tools	75/100 (3/4)	75/100 (3/4)	no change
feature-flags/flag-and-release-change	-	100/100 (4/4)	new
feature-flags/flag-release	-	100/100 (5/5)	new
feature-flags/launchdarkly-flag-command	-	100/100 (3/3)	new
feature-flags/launchdarkly-flag-create	100/100 (3/3)	75/100 (3/4)	-25
feature-flags/launchdarkly-flag-drift	-	100/100 (4/4)	new
feature-flags/should-flag-change	-	100/100 (10/10)	new

Only suites whose source actually changed since their last recorded score were re-run. Soft-failing while we stabilise the baseline.

[darklauncher-noesquick]

+1 to @erangeles , I'd love to see the video 😄

[darklauncher-noesquick]

Lgtm though!

[fsaresh]

@fsaresh can you please add a video or recording of the skill for context please?

Here's a demo I might re-record / might not be TTW-worthy just yet 🙂

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 7f33a56. Configure here.}