Install
$ agentstack add skill-xeldaralz-everything-claude-unity-deep-interview ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Deep Interview — Ambiguity Gating
Before executing any feature request, evaluate whether the requirements are clear enough to proceed. Vague requests waste agent cycles and produce code that needs reworking.
When to Activate
A request is vague when ALL of these are true:
- No file paths or directory references mentioned
- No function, class, or component names specified
- No code blocks or snippets included
- No numbered steps or explicit instructions
- The request is more than 10 words (short requests like "add a jump" are intentionally open-ended and acceptable)
If any one specificity signal is present, skip the interview and proceed normally.
Exemptions — Skip the Interview
Do NOT activate for:
- Bug fixes — request mentions "error", "bug", "crash", "fix", "broken", "NullReference", "exception"
- Commands with built-in clarification —
/unity-workflowhas its own Phase 1: Clarify,/unity-prototypeexpects open-ended prompts - Explicit opt-out — user includes
--skip-interviewor says "just do it" / "don't ask" - Follow-up requests — user is continuing a conversation where requirements were already established
- Simple modifications — "make it faster", "change the color to red", "increase the speed"
Ambiguity Score
Rate the request across 5 dimensions. Each scores 0–2:
| Score | Meaning | |-------|---------| | 0 | Unspecified — no information provided | | 1 | Partial — vague or implied | | 2 | Clear — explicitly stated or obvious from context |
Dimensions
- Scope — What exactly is being built? What are its boundaries? What is NOT included?
- Platform — Target platform, Unity version, render pipeline, input method?
- Performance — FPS target, memory budget, draw call limits, target device tier?
- Integration — What existing systems does this touch? Dependencies? Data flow?
- Acceptance Criteria — How do we know it's done? What should we test? What does success look like?
Threshold: total score >= 6 out of 10 to proceed.
Interview Protocol
When the score is below threshold:
- Present the current scores — show the user which dimensions are weak
- Ask targeted questions — max 3 questions per round, focused on the lowest-scoring dimensions
- Re-score after each round — update scores based on answers
- Proceed when threshold is met — or when the user explicitly opts out
Question style: be direct and specific, not generic. Instead of "What platform?", ask "Is this for mobile (touch input, limited GPU) or desktop (keyboard/mouse, full GPU budget)?"
Output: Requirements Document
When the threshold is met, produce a structured summary:
## Requirements Summary
**Feature:** [one-line description]
**Scope:** [what's included and what's explicitly excluded]
**Platform:** [target platform, input method, render pipeline]
**Performance:** [constraints, budgets, target devices]
**Integration:** [systems touched, dependencies, data flow]
**Acceptance Criteria:**
- [criterion 1]
- [criterion 2]
- [criterion 3]
Ask the user to confirm the summary before proceeding to implementation.
Scoring Examples
Vague (score 3/10): "Add multiplayer to my game"
- Scope: 1 (multiplayer is broad — co-op? competitive? matchmaking?)
- Platform: 0 (unspecified)
- Performance: 0 (unspecified)
- Integration: 1 (implies networking but no specifics)
- Acceptance: 1 (implied: "it works")
Clear enough (score 7/10): "Add 2-player local co-op split-screen for the existing PlayerController using the new Input System"
- Scope: 2 (2-player local co-op split-screen)
- Platform: 1 (implies desktop from split-screen, but not explicit)
- Performance: 1 (split-screen implies rendering budget concern)
- Integration: 2 (PlayerController, Input System explicitly named)
- Acceptance: 1 (implied: both players can play simultaneously)
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: XeldarAlz
- Source: XeldarAlz/everything-claude-unity
- License: MIT
- Homepage: https://github.com/XeldarAlz/everything-claude-unity#readme
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.