Anti Hallucination logo

Anti Hallucination

Community
Aedelon
anti-hallucination

CRITICAL SAFETY SKILL — Verify ALL technical claims, API signatures, library methods, code behavior, and factual statements before answering. Use Context7, WebSearch, and Read tools to confirm. MUST BE USED PROACTIVELY whenever: writing code that calls any library or API, answering technical questions, stating facts about frameworks or tools, mentioning version numbers, describing function behavior, recommending libraries, comparing technologies, or making any claim that could be wrong. Triggers: "how does X work", "what are the arguments for", "does this function exist", "is this the right syntax", "what version", "API reference", "documentation for", "how to use", "method signature", "return type", "parameters", "correct usage", "does this library support", "what's the default", "is this deprecated". Prevents hallucinated code, wrong function names, fabricated documentation, and incorrect facts. When in doubt about ANY technical claim, this skill MUST activate.

Overview

PublisherAedelon
Repositoryclaude-code-blueprint
Skill nameanti-hallucination
Stars
119
Forks
34
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by Aedelon on GitHub. Read the source before you install it.

Installation

Install the Anti Hallucination AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/Aedelon/claude-code-blueprint/tree/main/skills/anti-hallucination
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/Aedelon/claude-code-blueprint.git /tmp/claude-code-blueprint
mkdir -p .claude/skills
cp -r /tmp/claude-code-blueprint/skills/anti-hallucination .claude/skills/anti-hallucination
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Anti Hallucination in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Anti Hallucination on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Anti Hallucination is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Anti-Hallucination Protocol

Decision Tree (MANDATORY)

Question type?
├── API/Library signature → Context7 FIRST, THEN answer
├── Recent event/fact (< 1 year) → WebSearch FIRST
├── File content → Read tool FIRST
├── Code behavior → Read + trace FIRST
├── Historical fact → Can use training data
└── Cannot verify → State "I don't know"

Forbidden Actions (NEVER)

  • NEVER invent function signatures
  • NEVER guess library versions
  • NEVER assume API behavior without docs
  • NEVER fabricate citations or URLs
  • NEVER claim certainty without verification
  • NEVER make up statistics or numbers
  • NEVER invent paper titles or authors

Confidence Declaration

LevelCriteriaResponse Format
HIGHVerified via tool, 2+ sources"According to [source]: ..."
MEDIUMSingle reliable source"Based on [source], but should verify: ..."
LOWMemory only, no verification"I believe that... but I need to verify"
UNKNOWNNo data available"I don't know, would you like me to search?"

Verification Workflow

Step 1: Identify Claim Type

Claim TypeVerification Tool
Library APImcp__context7__get-library-docs
General factWebSearch
Specific URLmcp__fetch__fetch or WebFetch
File contentRead
Code patternGrep

Step 2: Execute Verification

BEFORE answering:

  1. Use appropriate tool
  2. Read the result carefully
  3. Extract relevant information
  4. Note any version constraints

Step 3: Cite Source in Response

# Good response
According to React 18.2 documentation: `useEffect` accepts two arguments...
Source: Context7 /facebook/react

# Bad response
useEffect accepts two arguments...  (no citation)

High-Risk Areas (Extra Caution)

API Signatures

HIGH RISK: Method names, parameter order, return types
ACTION: Always verify with Context7 before stating

Example:
❌ "The function takes (a, b, c) as parameters"
✅ "According to Context7: fetch(url, options?) → Promise<Response>"

Version-Specific Behavior

HIGH RISK: Breaking changes between versions
ACTION: State version explicitly

Example:
❌ "Next.js uses the App Router"
✅ "Next.js 13+ uses App Router (Pages Router before)"

Recent Changes

HIGH RISK: Features added/removed recently
ACTION: WebSearch for confirmation

Example:
❌ "React 19 introduces..."
✅ "According to web search (Dec 2024): React 19..."

Response Templates

Verified Answer

[Response based on verification]

**Source**: [Tool used] - [Source detail]
**Confidence**: HIGH

Partially Verified

[Response with some parts verified]

**Verified**: [What was confirmed]
**Unverified**: [What remains to confirm]
**Confidence**: MEDIUM

Cannot Verify

I don't have reliable information on this point.

**Options**:
1. I can search using [appropriate tool]
2. Consult [suggested source]
3. [Alternative if applicable]

Self-Check Before Response

□ Did I verify API signatures?
□ Are versions explicit?
□ Are sources cited?
□ Is my confidence level declared?
□ Did I avoid inventing details?

Frequently asked questions

What does the Anti Hallucination AI skill do?

CRITICAL SAFETY SKILL — Verify ALL technical claims, API signatures, library methods, code behavior, and factual statements before answering. Use Context7, WebSearch, and Read tools to confirm. MUST BE USED PROACTIVELY whenever: writing code that calls any library or API, answering technical questions, stating facts about frameworks or tools, mentioning version numbers, describing function behavior, recommending libraries, comparing technologies, or making any claim that could be wrong. Triggers: "how does X work", "what are the arguments for", "does this function exist", "is this the right...

Why use Anti Hallucination on TypingMind?

Because you install it once and use it with any model. Anti Hallucination is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Anti Hallucination in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/Aedelon/claude-code-blueprint/tree/main/skills/anti-hallucination. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Anti Hallucination?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Anti Hallucination?

As many as you like. As long as a model supports skills, you can use Anti Hallucination with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Anti Hallucination AI skill free?

It is published on GitHub by Aedelon. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇