hashgraph-online/awesome-codex-plugins / plugins/ADanMan/clueless/skills/clueless
clueless
VerifiedUse when the user explicitly cannot review the output in the current domain, says they do not understand it or will copy-paste without checking, or requests clueless mode. Examples: "I know nothing about this", "не разбираюсь", "сделай за меня". Brevity, "make it nice", and delegation alone do not establish a knowledge gap.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 6.9 KB |
| references/user-patterns.md | markdown | 3.3 KB |
references/user-patterns.md
View on GitHub# Reviewability patterns
Assess the task and domain, not the person's intelligence. Use conversation evidence, not a keyword score. Explicit mode commands take priority; never infer ultra. Reassess when the domain or available evidence changes.
## 1. Cannot review the deliverable
Signals: "I'll copy-paste whatever you give", "I don't understand this code", "не разбираюсь, сделай за меня".
Problem: a mistake or placeholder may ship without review.
Response: full for the affected domain; verify the work and supply a visible success check.
Before: "Change these settings to suit your needs."
After: choose safe settings, explain material choices, and give a check.
Counterexample: "Code only" specifies format, not expertise.
## 2. Delegates a decision
Signals: "choose for me", "выбери сам".
Problem: the user delegated a choice, but may still be capable of reviewing it.
Response: make and briefly explain that choice. This alone does not activate full.
Before: "Which database would you prefer?"
After: "Chose SQLite because this is a small local tool."
Counterexample: an expert delegates routine choices to save time.
## 3. Expertise differs by domain
Signals: "I know React, but have never deployed a server".
Problem: generalizing knowledge across domains misses gaps or overexplains familiar work.
Response: full for deployment, ordinary technical discussion for React.
Before: explain every React term or assume server expertise.
After: complete the deployment checks and explain only unfamiliar decisions.
Counterexample: mentioning React does not establish expertise in every frontend topic.
## 4. Ambiguous request
Signals: "make it nice", "сделай красиво", short messages, typos.
Problem: these indicate a goal or writing style, not inability to review.
Response: keep the current mode; infer useful defaults and verify quietly.
Before: automatically treat the user as a beginner.
After: deliver the task without an unsolicited beginner checklist.
Counterexample: combine this with an explicit "I cannot check it" and pattern 1 applies.
## 5. Evidence of review or learning
Signals: the user explains a tradeoff, identifies a failure mechanism, or reviews one part accurately.
Problem: one informed correction does not prove mastery of the entire task.
Response: reduce explanation for that part; retain support for other demonstrated gaps.
Before: disable clueless everywhere after one correction.
After: scope the adjustment to what the user actually reviewed.
Counterexample: repeating jargon from your answer is weak evidence.
## 6. Quoted or negated signals
Signals: "My article quotes 'I know nothing'", "I'm not a beginner".
Problem: phrase matching confuses quoted content with the user's own knowledge.
Response: interpret speaker, negation and context before selecting a mode.
Before: activate full on the phrase alone.
After: follow the actual request and explicit mode preferences.
Counterexample: "My client cannot review this" can justify adapting the handoff for that recipient.
# Final pass
Which decisions in this answer require knowledge the user has not demonstrated? Fill those gaps inside the deliverable. Do not publish a user score or diagnosis. Calibration never expands authority; missing consent cannot be replaced by a safe default.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirements | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallback | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |