Back to skills
SkillHub ClubAnalyze Data & AIFull StackData / AI

ai-collaboration-standards

Prevent AI hallucination and ensure evidence-based responses when analyzing code or making suggestions. Use when: analyzing code, making recommendations, providing options, or when user asks about confidence/certainty. Keywords: certainty, assumption, inference, evidence, source, 確定性, 推測, 假設, 來源, 證據.

Packaged view

This page reorganizes the original catalog entry around fit, installability, and workflow context first. The original raw source lives below.

Stars
44
Hot score
91
Updated
March 20, 2026
Overall rating
C2.7
Composite score
2.7
Best-practice grade
A92.4

Install command

npx @skill-hub/cli install asiaostrich-universal-dev-skills-ai-collaboration-standards

Repository

AsiaOstrich/universal-dev-skills

Skill path: skills/ai-collaboration-standards

Prevent AI hallucination and ensure evidence-based responses when analyzing code or making suggestions. Use when: analyzing code, making recommendations, providing options, or when user asks about confidence/certainty. Keywords: certainty, assumption, inference, evidence, source, 確定性, 推測, 假設, 來源, 證據.

Open repository

Best for

Primary workflow: Analyze Data & AI.

Technical facets: Full Stack, Data / AI.

Target audience: everyone.

License: Unknown.

Original source

Catalog source: SkillHub Club.

Repository owner: AsiaOstrich.

This is still a mirrored public skill entry. Review the repository before installing into production workflows.

What it helps with

  • Install ai-collaboration-standards into Claude Code, Codex CLI, Gemini CLI, or OpenCode workflows
  • Review https://github.com/AsiaOstrich/universal-dev-skills before adding ai-collaboration-standards to shared team environments
  • Use ai-collaboration-standards for development workflows

Works across

Claude CodeCodex CLIGemini CLIOpenCode

Favorites: 0.

Sub-skills: 0.

Aggregator: No.

Original source / Raw SKILL.md

---
name: ai-collaboration-standards
description: |
  Prevent AI hallucination and ensure evidence-based responses when analyzing code or making suggestions.
  Use when: analyzing code, making recommendations, providing options, or when user asks about confidence/certainty.
  Keywords: certainty, assumption, inference, evidence, source, 確定性, 推測, 假設, 來源, 證據.
---

# AI Collaboration Standards

This skill ensures AI assistants provide accurate, evidence-based responses without hallucination.

## Quick Reference

### Certainty Tags

| Tag | Use When |
|-----|----------|
| `[Confirmed]` / `[已確認]` | Direct evidence from code/docs |
| `[Inferred]` / `[推論]` | Logical deduction from evidence |
| `[Assumption]` / `[假設]` | Based on common patterns (needs verification) |
| `[Unknown]` / `[未知]` | Information not available |
| `[Need Confirmation]` / `[待確認]` | Requires user clarification |

### Source Types

| Source Type | Tag | Reliability |
|-------------|-----|-------------|
| Project Code | `[Source: Code]` | ⭐⭐⭐⭐⭐ Highest |
| Project Docs | `[Source: Docs]` | ⭐⭐⭐⭐ High |
| External Docs | `[Source: External]` | ⭐⭐⭐⭐ High |
| Web Search | `[Source: Search]` | ⭐⭐⭐ Medium |
| AI Knowledge | `[Source: Knowledge]` | ⭐⭐ Low |
| User Provided | `[Source: User]` | ⭐⭐⭐ Medium |

### Core Rules

1. **Evidence-Based Only**: Only analyze content that has been explicitly read
2. **Cite Sources**: Include file path and line number for code references
3. **Classify Certainty**: Tag all statements with certainty level
4. **Always Recommend**: When presenting options, include a recommended choice with reasoning

## Detailed Guidelines

For complete standards, see:
- [Anti-Hallucination Guidelines](./anti-hallucination.md)
- [Certainty Labels Reference](./certainty-labels.md)

## Examples

### ✅ Correct Response

```
[Confirmed] src/auth/service.ts:45 - JWT validation uses 'jsonwebtoken' library
[Inferred] Based on repository pattern in src/repositories/, likely using dependency injection
[Need Confirmation] Should the new feature support multi-tenancy?
```

### ❌ Incorrect Response

```
The system uses Redis for caching (code not reviewed)
The UserService should have an authenticate() method (API not verified)
```

### ✅ Correct Option Presentation

```
There are three options:
1. Redis caching
2. In-memory caching
3. File-based caching

**Recommended: Option 1 (Redis)**: Given the project already has Redis infrastructure
and needs cross-instance cache sharing, Redis is the most suitable choice.
```

### ❌ Incorrect Option Presentation

```
There are three options:
1. Redis caching
2. In-memory caching
3. File-based caching

Please choose one.
```

## Checklist

Before making any statement:

- [ ] Source Verified - Have I read the actual file/document?
- [ ] Source Type Tagged - Did I specify `[Source: Code]`, `[Source: External]`, etc.?
- [ ] Reference Cited - Did I include file path and line number?
- [ ] Certainty Classified - Did I tag as `[Confirmed]`, `[Inferred]`, etc.?
- [ ] No Fabrication - Did I avoid inventing APIs, configs, or requirements?
- [ ] Recommendation Included - When presenting options, did I include a recommended choice?

---

## Configuration Detection

This skill supports project-specific language configuration for certainty tags.

### Detection Order

1. Check `CONTRIBUTING.md` for "Certainty Tag Language" section
2. If found, use the specified language (English / 中文)
3. If not found, **default to English** tags

### First-Time Setup

If no configuration found and context is unclear:

1. Ask the user: "This project hasn't configured certainty tag language preference. Which would you like to use? (English / 中文)"
2. After user selection, suggest documenting in `CONTRIBUTING.md`:

```markdown
## Certainty Tag Language

This project uses **[English / 中文]** certainty tags.
<!-- Options: English | 中文 -->
```

### Configuration Example

In project's `CONTRIBUTING.md`:

```markdown
## Certainty Tag Language

This project uses **English** certainty tags.

### Tag Reference
- [Confirmed] - Direct evidence from code/docs
- [Inferred] - Logical deduction from evidence
- [Assumption] - Based on common patterns
- [Unknown] - Information not available
- [Need Confirmation] - Requires user clarification
```

---

**License**: CC BY 4.0 | **Source**: [universal-dev-standards](https://github.com/AsiaOstrich/universal-dev-standards)


---

## Referenced Files

> The following files are referenced in this skill and included for context.

### anti-hallucination.md

```markdown
# Anti-Hallucination Guidelines

## Core Principles

### 1. Evidence-Based Analysis Only

**Rule**: Only analyze and reference content that has been explicitly provided or read.

**Guidelines**:
- ✅ Analyze code files that have been read using file reading tools
- ✅ Reference documentation that has been fetched
- ✅ Cite configuration files that have been inspected
- ❌ Do NOT speculate about APIs, functions, or configurations not seen
- ❌ Do NOT assume framework behavior without verification
- ❌ Do NOT fabricate requirement details

---

### 2. Explicit Source Attribution

**Rule**: All references must include source type, location, and verifiability information.

#### Source Types

| Source Type | Tag | Reliability |
|-------------|-----|-------------|
| Project Code | `[Source: Code]` | ⭐⭐⭐⭐⭐ Highest |
| Project Docs | `[Source: Docs]` | ⭐⭐⭐⭐ High |
| External Docs | `[Source: External]` | ⭐⭐⭐⭐ High |
| Web Search | `[Source: Search]` | ⭐⭐⭐ Medium |
| AI Knowledge | `[Source: Knowledge]` | ⭐⭐ Low |
| User Provided | `[Source: User]` | ⭐⭐⭐ Medium |

#### Attribution Format

**For Code References**:
```
[Source: Code] file_path:line_number - Description
```

**For External Documentation**:
```
[Source: External] URL - Description (Version: x.x.x, Accessed: YYYY-MM-DD)
```

**For AI Knowledge**:
```
[Source: Knowledge] Topic - Description (⚠️ Requires verification)
```

#### Version Sensitivity

When referencing libraries, frameworks, or APIs, always include version information:

```
✅ [Source: External] Next.js App Router (v14.x) - Server Components are the default
✅ [Source: Code] package.json:12 - Using "express": "^4.18.2"
❌ "Next.js uses Server Components" (which version?)
```

---

### 3. Classify Certainty Levels

**Rule**: Clearly distinguish between confirmed facts, inferences, and unknowns.

| Tag | Use When |
|-----|----------|
| `[Confirmed]` | Direct evidence from code/docs |
| `[Inferred]` | Logical deduction from available evidence |
| `[Assumption]` | Based on common patterns (needs verification) |
| `[Unknown]` | Information not available |
| `[Need Confirmation]` | Requires user clarification |

**Examples**:

```
[Confirmed] src/database/connection.ts:12 - Using PostgreSQL driver 'pg'
[Inferred] Based on the repository pattern in src/repositories/, likely using dependency injection
[Assumption] Project may use OAuth2, but need to review auth configuration
[Unknown] API rate limiting strategy not documented
[Need Confirmation] Should the new feature support multi-tenancy?
```

---

### 4. Recommendation Principles

**Rule**: When providing multiple options, always include a recommended choice with reasoning.

| Scenario | Requirement |
|----------|-------------|
| 2+ options presented | Must indicate recommended option with reasoning |
| Clear winner exists | Directly recommend best option with reasoning |
| Trade-offs exist | Recommend based on current context, explain trade-offs |
| Cannot determine | Explain what information is needed to make a recommendation |

---

### 5. Prohibited Behaviors

AI assistants MUST NOT:

1. **Fabricate APIs or Function Signatures**
   - ❌ Do NOT invent method names, parameters, or return types
   - ✅ DO read the actual source code or ask the user

2. **Assume Requirements**
   - ❌ Do NOT guess user needs or business rules
   - ✅ DO ask clarifying questions when requirements are ambiguous

3. **Speculate About Unread Code**
   - ❌ Do NOT describe functionality of files not reviewed
   - ✅ DO explicitly state "Need to read [file] to confirm"

4. **Invent Configuration**
   - ❌ Do NOT assume environment variables, config keys, or database schemas
   - ✅ DO review actual configuration files

5. **Hallucinate Errors or Bugs**
   - ❌ Do NOT claim code has issues without evidence
   - ✅ DO analyze actual code and cite specific lines

6. **Present Options Without Recommendation**
   - ❌ Do NOT list options and ask user to choose without guidance
   - ✅ DO always include a recommended choice with reasoning

---

## AI Assistant Workflow

```
┌─────────────────────────────────┐
│  User Request Received          │
└─────────────┬───────────────────┘
              │
              ▼
┌─────────────────────────────────┐
│  Identify Information Needed    │
│  - Code files?                  │
│  - Configuration?               │
│  - Requirements?                │
└─────────────┬───────────────────┘
              │
              ▼
         ┌────┴────┐
         │ Available? │
         └────┬────┘
              │
      ┌───────┴───────┐
      │               │
     YES              NO
      │               │
      ▼               ▼
┌──────────┐   ┌─────────────┐
│  Read/   │   │  Ask User   │
│  Analyze │   │  for Info   │
└────┬─────┘   └──────┬──────┘
     │                │
     ▼                ▼
┌─────────────────────────────────┐
│  Tag Response with:             │
│  - [Confirmed] for facts        │
│  - [Inferred] for deductions    │
│  - [Need Confirmation] for gaps │
└─────────────┬───────────────────┘
              │
              ▼
┌─────────────────────────────────┐
│  Cite Sources (file:line)       │
└─────────────┬───────────────────┘
              │
              ▼
┌─────────────────────────────────┐
│  Include Recommendation         │
│  (if presenting options)        │
└─────────────────────────────────┘
```

---

**License**: CC BY 4.0 | **Source**: [universal-dev-standards](https://github.com/AsiaOstrich/universal-dev-standards)

```

### certainty-labels.md

```markdown
# Certainty Labels Reference

## Tag Mapping (English / 中文)

| English Tag | 中文標籤 | Use When |
|-------------|---------|----------|
| `[Confirmed]` | `[已確認]` | Direct evidence from code/docs |
| `[Inferred]` | `[推論]` | Logical deduction from evidence |
| `[Assumption]` | `[假設]` | Based on common patterns (needs verification) |
| `[Unknown]` | `[未知]` | Information not available |
| `[Need Confirmation]` | `[待確認]` | Requires user clarification |

---

## Source Types

| Source Type | Tag | Description | Reliability |
|-------------|-----|-------------|-------------|
| Project Code | `[Source: Code]` | Directly read from codebase | ⭐⭐⭐⭐⭐ Highest |
| Project Docs | `[Source: Docs]` | README, Wiki, inline comments | ⭐⭐⭐⭐ High |
| External Docs | `[Source: External]` | Official documentation with URL | ⭐⭐⭐⭐ High |
| Web Search | `[Source: Search]` | Search results (include date) | ⭐⭐⭐ Medium |
| AI Knowledge | `[Source: Knowledge]` | AI training data (needs verification) | ⭐⭐ Low |
| User Provided | `[Source: User]` | Information from user conversation | ⭐⭐⭐ Medium |

---

## Usage Examples

### In Technical Documents

```markdown
## System Architecture Analysis

`[Confirmed]` System uses ASP.NET Core 8.0 framework [Source: Code] Program.cs:1
`[Confirmed]` Database uses SQL Server [Source: Code] appsettings.json:12
`[Inferred]` Based on Repository Pattern usage, system likely adopts DDD architecture
`[Assumption]` Caching mechanism may use Redis (need to confirm config)
`[Need Confirmation]` Should multi-tenancy be supported?
```

### 技術文件中(中文)

```markdown
## 系統架構分析

`[已確認]` 系統使用 ASP.NET Core 8.0 框架 [Source: Code] Program.cs:1
`[已確認]` 資料庫採用 SQL Server [Source: Code] appsettings.json:12
`[推論]` 基於 Repository Pattern 的使用,系統可能採用 DDD 架構
`[假設]` 快取機制可能使用 Redis(需確認設定檔)
`[待確認]` 是否需要支援多租戶架構?
```

### In Code Review

```markdown
## Review Comments

`[Confirmed]` src/Services/AuthService.cs:45 - Password validation lacks brute force protection
`[Inferred]` Rate limiting may be needed here
`[Need Confirmation]` Are there other layers of protection already in place?
```

### 程式碼審查中(中文)

```markdown
## 審查意見

`[已確認]` src/Services/AuthService.cs:45 - 密碼驗證缺少防暴力破解機制
`[推論]` 此處可能需要加入 Rate Limiting
`[待確認]` 是否已有其他層級的防護措施?
```

---

## Best Practices

### 1. Consistency

- Use the same language tags throughout a document (all Chinese or all English)
- Team should specify preferred language in `CONTRIBUTING.md`

### 2. Source Citation

- Chinese tags still require source citations
- Format: `[已確認]` statement [Source: Code] file_path:line_number

### 3. Team Agreement

- Decide on Chinese or English tags at project start
- Document in `CONTRIBUTING.md` or `.standards/` directory

---

## Quick Decision Guide

```
Did you read the actual code/doc?
├── YES → [Confirmed] / [已確認]
└── NO
    ├── Can you deduce from available evidence?
    │   ├── YES → [Inferred] / [推論]
    │   └── NO
    │       ├── Is it a common pattern?
    │       │   ├── YES → [Assumption] / [假設]
    │       │   └── NO → [Unknown] / [未知]
    └── Does user need to clarify?
        └── YES → [Need Confirmation] / [待確認]
```

---

**License**: CC BY 4.0 | **Source**: [universal-dev-standards](https://github.com/AsiaOstrich/universal-dev-standards)

```

ai-collaboration-standards | SkillHub