From compound-engineering
This skill should be used when the user asks to "assess progress", "find gaps", "what should I work on next", "push the project forward", "gap analysis", "unblock the project", or when deciding what work matters most toward a goal. Works for any domain — engineering, product, operations, content, strategy. Provides the thinking methodology for pragmatic gap identification and prioritization.
npx claudepluginhub mberto10/mberto-compoundThis skill uses the workspace's default tool permissions.
Identify what stands between the current state and a goal, then rank gaps by pragmatic impact. This is the **strategic targeting** lens — applied before doing work to decide *what* to work on, not *how* to do it.
Generates design tokens/docs from CSS/Tailwind/styled-components codebases, audits visual consistency across 10 dimensions, detects AI slop in UI.
Records polished WebM UI demo videos of web apps using Playwright with cursor overlay, natural pacing, and three-phase scripting. Activates for demo, walkthrough, screen recording, or tutorial requests.
Delivers idiomatic Kotlin patterns for null safety, immutability, sealed classes, coroutines, Flows, extensions, DSL builders, and Gradle DSL. Use when writing, reviewing, refactoring, or designing Kotlin code.
Identify what stands between the current state and a goal, then rank gaps by pragmatic impact. This is the strategic targeting lens — applied before doing work to decide what to work on, not how to do it.
Works for any goal: shipping a feature, launching a product, completing a migration, preparing a presentation, rolling out a process, hiring a team, or reaching a business target.
"If I could only do one thing to move this forward, what would it be?"
Everything in this skill serves that question.
When surveying state against a goal, classify every gap:
| Type | Meaning | Signal |
|---|---|---|
| MISSING | Hasn't been started, no work exists | No deliverables, no issues, no progress |
| INCOMPLETE | Started but not finished | Open tasks, partial drafts, work in progress |
| BROKEN | Exists but doesn't work correctly | Failing validation, known issues, negative feedback |
| BLOCKED | Can't progress without resolving a dependency | Waiting on external input, decision, approval, upstream work |
| UNKNOWN | Unclear state, needs investigation | No recent activity, ambiguous scope, nobody owns it |
UNKNOWN is the most dangerous type. It hides risk. Always recommend investigating UNKNOWNs before committing to large MISSING work.
Not all gaps are equal. Rank by these criteria, in order:
A gap that blocks three other things is more valuable to close than a gap that blocks nothing. Map the dependency chain:
Gap A blocks → Gap B, Gap C
Gap D blocks → nothing
→ Close Gap A first, even if Gap D is "easier"
Quick win (< 1 session, high impact) → Do immediately
Deep work (multi-session, high impact) → Plan and schedule
Low-hanging (< 1 session, low impact) → Batch with other work
Slog (multi-session, low impact) → Defer or cut scope
| Anti-Pattern | What's Happening | Correction |
|---|---|---|
| Polishing before foundations | Refining details when core deliverables are incomplete | Prioritize BROKEN and MISSING in critical paths first |
| Optimizing before validating | Improving efficiency when correctness isn't proven | Validate the approach works before making it elegant |
| Scope creep disguised as gaps | "We also need X" when X isn't in the goal | Check if the gap is actually in scope |
| Busywork over blockers | Doing easy INCOMPLETE work while BLOCKED items rot | Address BLOCKED items — they often need a decision, not more work |
| Analysis paralysis on UNKNOWN | Investigating endlessly instead of time-boxing | Set a time box: 30 min to classify, then decide |
| Comfort zone bias | Working on what's familiar while avoiding the hard gap | The uncomfortable gap is usually the important one |
When assessing progress, gather state from whatever sources are available. Not all sources apply to every goal — use what's relevant.
When the goal involves engineering work:
/plan the implementation/review the area first/explore-subsystemWhen the goal involves project management:
/draft-communication/prepare-updateIn all cases:
/discover/reflectWhen you need a fast assessment (< 5 minutes), use this:
Goal: {one line}
Current state: {one line}
Top 3 gaps:
1. {gap} [{type}] — blocks: {what} — effort: {quick/deep}
2. {gap} [{type}] — blocks: {what} — effort: {quick/deep}
3. {gap} [{type}] — blocks: {what} — effort: {quick/deep}
→ Do next: {#N because reason}