Cassiiopeia/projectops / skills/pro-skill-creator
pro-skill-creator
VerifiedCreate, review or improve a standard skill. Classifies intent as create/review/improve, then runs the matching phases. Use for 'make a skill', 'review this skill', 'improve this skill'.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 10.8 KB |
| references/anti_patterns.md | markdown | 4.4 KB |
| references/meta_principles.md | markdown | 5.1 KB |
| references/phase0_brainstorming.md | markdown | 8.6 KB |
| references/phase1_slots.md | markdown | 3.4 KB |
| references/phase2_question_format.md | markdown | 1.8 KB |
| references/phase4_checklist.md | markdown | 3.1 KB |
| references/phase6_report_format.md | markdown | 2.2 KB |
| references/review_improve.md | markdown | 4.0 KB |
| references/trigger_optimization.md | markdown | 2.6 KB |
| templates/config.json.example | other | 495 B |
| templates/python_cli_script.py | script | 3.9 KB |
references/phase4_checklist.md
View on GitHub# Phase 4 — 8-principle self-review checklist
Shared by CREATE mode Phase 4 and REVIEW mode Phase 2.
**Mandatory procedure**: **always re-read the file you just wrote with the `Read` tool** and fill in the 8 checks below one line each. "I just wrote it, so I know without looking" is not allowed.
## Checklist
| # | Principle | PASS condition | Fix on violation | Evidence (where to check) |
|---|------|----------|-------------|-----------------|
| 1 | One at a time | No place in the Phase 0/Phase 2 sections puts 2 or more questions in one message? | Split the questions | Question examples in the SKILL.md body |
| 2 | Automatic inference first | Not asking again for values inferable from user input (URL → domain, compose → port, etc.)? | Change to automatic extraction | Scripts + SKILL.md instructions |
| 3 | Explain technical terms | Each abbreviation/proper noun has a one-line explanation where it first appears? | Add a one-line explanation | Skim the whole body |
| 4 | Concrete UI paths | UI directions are at the "menu → tab → field name" level? No "check it there" | Add path detail | Search for vague words like "check", "settings" |
| 5 | No hard-coding | No real domain/IP/email/employee ID baked into the body? | Replace with `{placeholder}` | `grep` for `.com`, `@`, IP patterns |
| 6 | Separate config file | Values that differ per user moved into a config file? | Add a config file definition | Check that config.json.example exists |
| 7 | Reflects real experience | Warnings written as "general advice" were verified by user experience? Unverified warnings only confuse | Delete or mark `(unverified)` | Skim the pitfalls section |
| 8 | Flexible input | Is there a branch that accepts it when the user gives all the information at once? | Add an automatic extraction branch | Argument parsing logic in scripts |
## Extra checks from the official skill-creator best practices (REVIEW mode only)
On top of the 8 principles, REVIEW mode also checks the 4 items below.
| # | Criterion | PASS condition | Fix on violation |
|---|------|-----------|-------------|
| a | Lean prompt | Are there decorative sentences or duplicate MUSTs that could go without changing results? | Delete |
| b | Explain the why | Do "do not" rules come with a reason? | Attach one line `(reason: ~)` |
| c | Progressive disclosure | SKILL.md ≤ 500 lines? If over, is it split into references/? | Split |
| d | Bundle repeated work | Is repeated logic moved into scripts/? | Extract into a script |
## If even one item FAILs
Fix SKILL.md immediately and check again. Repeat until there are 0 violations. Never declare "roughly fine" to the user.
## Special cases
- **The skill runs in a network-restricted environment** → read the "automatic extraction" of principle 2 (automatic inference) as "must work without external API calls".
- **The skill wraps a single command** → the phase split (Phase 0/1/2) may be omitted. The 8 principles still apply.
- **Description language** → write the `description` in English by default. If the users speak another language (for example a mixed Korean/English team), also include trigger keywords in that language. A single-language team may use only one.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |