Knowledge Base logo

Knowledge Base

Organization
cosmicstack-labs
knowledge-base

Knowledge Base Creation: Searchable wikis, internal documentation portals, knowledge management systems

Overview

Publishercosmicstack-labs
Repositorymercury-agent-skills
Skill nameknowledge-base
Stars
485
Forks
63
Bundled files
Instructions only
LicenseMIT
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 cosmicstack-labs on GitHub. Read the source before you install it.

Installation

Install the Knowledge Base 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/cosmicstack-labs/mercury-agent-skills/tree/main/categories/development/knowledge-base
  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/cosmicstack-labs/mercury-agent-skills.git /tmp/mercury-agent-skills
mkdir -p .claude/skills
cp -r /tmp/mercury-agent-skills/categories/development/knowledge-base .claude/skills/knowledge-base
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Knowledge Base 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 Knowledge Base 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 Knowledge Base 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.

Knowledge Base Creation

Build searchable, maintainable knowledge bases that your team actually uses — turning tribal knowledge into organizational assets.

Core Principles

1. Knowledge Bases Are Products, Not Projects

A knowledge base is never "done." It requires ongoing investment: content creation, curation, cleanup, and promotion. Treat it like a product with a roadmap, owners, and metrics.

2. Search Is the Primary Interface

If users can't find what they need in 2-3 searches, the knowledge base has failed. Invest in search quality, metadata, tagging, and cross-referencing before adding more content.

3. Lower the Barrier to Contribution

The easier it is to write and publish, the more content you'll get. Support markdown, provide templates, reduce review friction for minor edits, and celebrate contributors.

4. Structure for Discoverability, Write for Scannability

Organize content in categories users naturally think in. Use clear titles, descriptive summaries, consistent formatting, and plenty of internal links.

5. Measure What Matters

Track search success rates, popular content, stale pages, and user feedback. Use data to decide what to improve, archive, or create next.


Knowledge Base Maturity Model

LevelContentSearchMaintenanceContributionGovernance
1: GraveyardRandom docs, no structureNone (browse only)Never updatedNo processNo owners
2: CollectionSome structure, inconsistentBasic text searchUpdated reactivelyA few contributorsUnclear ownership
3: OrganizedTaxonomy defined, templates usedTagged + full-text searchRegular reviewsActive contributorsDefined content owners
4: CuratedReviewed content, versionedFaceted + filtered searchScheduled auditsContributor workflowsEditorial board
5: LivingFeedback-driven, analytics-informedSemantic + AI-assistedContinuous improvementEmbedded in workflowsMetrics-driven governance

Target: Level 3 for most teams. Level 4 for customer-facing or compliance-relevant knowledge bases.


Actionable Guidance

Platform Selection

PlatformBest ForKey FeaturesLimitationsHosting
GitBookDeveloper docs, API docsGit-synced, versioned, search, multi-languagePricing at scaleCloud or self-hosted
NotionInternal wikis, team knowledgeRich editor, databases, templates, AI searchExport limitations, performanceCloud only
DocusaurusOpen source project docsReact-based, versioned, MDX, search pluginsRequires developer setupStatic site (any host)
MkDocsPython projects, simple docsPython ecosystem, Material theme, pluginsLess flexible than DocusaurusStatic site (any host)
ConfluenceEnterprise internal docsDeep Jira integration, permissions, analyticsHeavy, expensive, poor searchCloud or self-hosted
Obsidian PublishPersonal knowledge, ZettelkastenGraph view, backlinks, local-firstLimited collaborationCloud (paid)
OutlineInternal wikis (mid-size teams)Markdown-native, fast, Slack integrationSmaller ecosystemCloud or self-hosted
Decision Matrix
markdown
## Platform Decision Matrix

| Requirement | GitBook | Notion | Docusaurus | MkDocs | Confluence |
|-------------|---------|--------|------------|--------|------------|
| Version control (git) | ✅ Native | ❌ Manual | ✅ Native | ✅ Native | ❌ Manual |
| Developer-friendly | ✅ | ❌ | ✅✅ | ✅✅ | ❌ |
| Non-technical editors | ✅✅ | ✅✅ | ❌ | ❌ | ✅✅ |
| Self-hosted | ✅ | ❌ | ✅✅ | ✅✅ | ✅ |
| Advanced search | ✅ | ✅ | ✅ (plugin) | ✅ (plugin) | ❌ |
| API / integrations | ✅ | ✅ | ✅ (open source) | ✅ (open source) | ✅ |
| Cost (low) | ❌ (paid) | ❌ (paid) | ✅ (free) | ✅ (free) | ❌ (expensive) |
| Mobile app | ✅ | ✅ | ❌ | ❌ | ✅ |
Setup Comparison

Docusaurus (recommended for developer docs):

bash
# Quick start
npx create-docusaurus@latest my-docs classic
cd my-docs

# Add search (Algolia DocSearch)
npm install --save @docusaurus/plugin-search-algolia

# Add versioning
npm run docusaurus docs:version 1.0.0

# Build and deploy
npm run build
# Deploy `build/` to any static host
javascript
// docusaurus.config.js
module.exports = {
  title: 'My Project Docs',
  tagline: 'Everything you need to build with My Project',
  url: 'https://docs.myproject.com',
  baseUrl: '/',
  onBrokenLinks: 'throw',
  onBrokenMarkdownLinks: 'warn',
  favicon: 'img/favicon.ico',
  
  presets: [
    [
      'classic',
      {
        docs: {
          sidebarPath: require.resolve('./sidebars.js'),
          editUrl: 'https://github.com/myorg/myproject/edit/main/docs/',
          showLastUpdateTime: true,
          showLastUpdateAuthor: true,
        },
        blog: {
          showReadingTime: true,
        },
        theme: {
          customCss: require.resolve('./src/css/custom.css'),
        },
      },
    ],
  ],
  
  themeConfig: {
    navbar: {
      title: 'My Project',
      items: [
        {
          type: 'doc',
          docId: 'intro',
          position: 'left',
          label: 'Docs',
        },
        {
          type: 'docsVersionDropdown',
          position: 'right',
        },
        {
          href: 'https://github.com/myorg/myproject',
          label: 'GitHub',
          position: 'right',
        },
      ],
    },
    algolia: {
      appId: 'YOUR_APP_ID',
      apiKey: 'YOUR_API_KEY',
      indexName: 'my-project',
    },
  },
};

MkDocs with Material theme:

bash
# Install
pip install mkdocs mkdocs-material

# Create new project
mkdocs new .

# Serve locally
mkdocs serve

# Build
mkdocs build
yaml
# mkdocs.yml
site_name: My Project Docs
site_url: https://docs.myproject.com
theme:
  name: material
  features:
    - navigation.tabs
    - navigation.sections
    - navigation.expand
    - search.highlight
    - content.code.copy
    - content.tabs.link

plugins:
  - search
  - tags
  - git-revision-date-localized

markdown_extensions:
  - admonition
  - pymdownx.details
  - pymdownx.superfences
  - pymdownx.tabbed:
      alternate_style: true
  - tables
  - toc:
      permalink: true

nav:
  - Home: index.md
  - Getting Started: getting-started.md
  - Guides:
    - guides/index.md
    - Configuration: guides/configuration.md
    - Deployment: guides/deployment.md
  - Reference:
    - reference/index.md
    - CLI: reference/cli.md
    - API: reference/api.md
  - Troubleshooting: troubleshooting.md

Content Structuring

The Information Architecture Pyramid
text
                    ┌─────────────┐
                    │  Home/Landing │
                    │  (Overview)   │
                    └──────┬──────┘
           ┌───────────────┼───────────────┐
           │               │               │
    ┌──────┴──────┐ ┌─────┴──────┐ ┌──────┴──────┐
    │  Getting     │ │  Guides &  │ │  Reference   │
    │  Started     │ │  Tutorials │ │  & API Docs  │
    └──────┬──────┘ └─────┬──────┘ └──────┬──────┘
           │              │               │
    ┌──────┴──────┐ ┌─────┴──────┐ ┌──────┴──────┐
    │  Quick Start │ │  How-to    │ │  Config      │
    │  Setup Guide │ │  Best      │ │  CLI Ref     │
    │  First Steps │ │  Practices │ │  Changelog   │
    └─────────────┘ └────────────┘ └─────────────┘
Folder Structure Template
text
docs/
├── index.md                    # Landing page
├── getting-started/
│   ├── index.md               # Overview
│   ├── installation.md        # Setup instructions
│   ├── quick-start.md         # 5-minute tutorial
│   └── first-project.md       # Guided walkthrough
├── guides/
│   ├── index.md               # Guide overview
│   ├── configuration/
│   │   ├── environment.md
│   │   ├── authentication.md
│   │   └── integrations.md
│   ├── deployment/
│   │   ├── production.md
│   │   ├── staging.md
│   │   └── monitoring.md
│   └── best-practices/
│       ├── security.md
│       ├── performance.md
│       └── scalability.md
├── reference/
│   ├── index.md               # Reference overview
│   ├── api.md                 # API documentation
│   ├── cli.md                 # CLI commands
│   ├── configuration.md       # All config options
│   └── troubleshooting.md     # Common issues & solutions
├── tutorials/
│   ├── index.md
│   ├── beginner-tutorial.md
│   └── advanced-tutorial.md
├── contributing.md            # How to contribute to the KB
└── faq.md                     # Frequently asked questions
Page Template
markdown
---
title: How to [Task Name]
description: Step-by-step guide for completing [task]
sidebar_position: 2
tags: [configuration, deployment, production]
---

# How to [Task Name]

> **Purpose**: [One-line description of what this accomplishes]
> **Prerequisites**: [List of prerequisites or links to them]
> **Estimated time**: [X minutes]

## Step 1: [First Step]

[Description of what to do]

```bash
[Command to run]

Expected result: [What should happen after this step]

Step 2: [Second Step]

[Description]

Verification

[How to confirm success]

Troubleshooting

ProblemSolution
[Problem description][Solution steps]

Related


---

### Taxonomy and Tagging

#### Tag Taxonomy Design

```yaml
# Tag taxonomy for a developer knowledge base
#
# Category-based tags follow this hierarchy:
# domain > area > topic

tags:
  # By topic area
  - backend
  - frontend
  - infrastructure
  - security
  - data
  - mobile
  
  # By content type
  - tutorial
  - guide
  - reference
  - troubleshooting
  - faq
  
  # By audience
  - beginner
  - intermediate
  - advanced
  - administrator
  
  # By system/component
  - authentication
  - database
  - deployment
  - monitoring
  - api
  - cli
  
  # By project/team
  - team-alpha
  - team-beta
  - platform
Tagging Guidelines
markdown
## Tagging Guidelines

### Rules
1. **Every page needs at least 2 tags**: one content type + one topic
2. **Max 5 tags per page**: too many tags dilute search
3. **Use existing tags**: check existing tags before creating new ones
4. **Keep tags lowercase**: consistent casing improves search
5. **Use nouns, not verbs**: `deployment` not `deploying`

### Tag Combinations
| Page Type | Required Tags | Recommended Tags |
|-----------|--------------|------------------|
| Getting started guide | `tutorial`, `beginner` | `setup`, `first-steps` |
| API reference | `reference`, `api` | `backend`, `integration` |
| Troubleshooting guide | `troubleshooting` | Component-related tags |
| Best practices | `guide`, `best-practices` | Domain-specific tags |
Automated Tag Suggestions
python
# Example: suggest tags based on page content
def suggest_tags(content: str, existing_tags: list) -> list:
    """
    Suggest relevant tags based on page content analysis.
    """
    keywords = {
        "install": ["getting-started", "setup"],
        "config": ["configuration", "setup"],
        "deploy": ["deployment", "devops"],
        "error": ["troubleshooting", "errors"],
        "api": ["api", "integration"],
        "security": ["security", "authentication"],
        "database": ["database", "data"],
        "monitor": ["monitoring", "observability"],
    }
    
    suggestions = set()
    content_lower = content.lower()
    
    for keyword, tags in keywords.items():
        if keyword in content_lower:
            suggestions.update(tags)
    
    # Limit to most relevant
    return list(suggestions)[:3]

Search Optimization

Content That Ranks in Internal Search
markdown
## Writing for Search Discovery

### 1. Start with the Question
Use the question as the title so it matches what users search for:

✅ "How do I reset my password?"
✅ "Why is my deployment failing with error 503?"
❌ "Password Reset Procedure"
❌ "Deployment Error Analysis"

### 2. Include Synonyms in the First Paragraph
Users search with different vocabulary:

> "If your login fails or you can't sign in to the dashboard,
> you may need to reset your password. This guide covers
> password recovery, credential reset, and account access
> restoration."

### 3. Use Descriptive Headings
Headings are heavily weighted in search:

✅ "### Resolving Database Connection Timeouts"
❌ "### Issue Resolution"

### 4. Add a Summary Block
```markdown
> **TL;DR**: If you see "Error 503 Service Unavailable" during
> deployment, your application server is overloaded. Scale up
> your instance or check for memory leaks before redeploying.

#### Search Configuration

```yaml
# MkDocs Material search configuration
plugins:
  - search:
      lang:
        - en
      separator: '[\s\-_]'
      min_search_length: 3
      prebuild_index: true  # Offline search support

# Docusaurus search (Algolia)
themeConfig:
  algolia:
    appId: YOUR_APP_ID
    apiKey: YOUR_API_KEY
    indexName: your-docs
    contextualSearch: true
    searchParameters:
      hitsPerPage: 10
      attributesToSnippet:
        - 'title:30'
        - 'content:200'

Knowledge Base Maintenance

Content Health Dashboard
markdown
## Content Health Metrics

Track these monthly:

| Metric | Target | How to Measure |
|--------|--------|---------------|
| Stale pages (>6 months since update) | <10% | Git log / last modified dates |
| Orphaned pages (no inbound links) | <5% | Backlink analysis |
| Search failure rate | <15% | Search analytics |
| Pages with no tags | <2% | Tag metadata audit |
| Reader satisfaction | >4/5 | Feedback widget |
| Pages per contributor (monthly) | >2 | Contribution tracking |
Stale Content Detection
yaml
# GitHub Action: Find stale docs
name: Stale Content Check
on:
  schedule:
    - cron: '0 6 1 * *'  # First day of every month

jobs:
  stale-check:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
        with:
          fetch-depth: 0
      
      - name: Find stale pages
        run: |
          echo "## Stale Documentation (not updated in 6 months)" > stale-report.md
          echo "" >> stale-report.md
          
          for file in $(find docs -name "*.md" -type f); do
            last_commit=$(git log -1 --format="%cd" --date=short -- "$file")
            if [[ $(date -d "$last_commit" +%s) -lt $(date -d "6 months ago" +%s) ]]; then
              echo "- [$file]($file) — last updated $last_commit" >> stale-report.md
            fi
          done
      
      - name: Create Issue
        uses: actions/github-script@v6
        with:
          script: |
            const fs = require('fs');
            const body = fs.readFileSync('stale-report.md', 'utf8');
            await github.rest.issues.create({
              owner: context.repo.owner,
              repo: context.repo.repo,
              title: 'Monthly Stale Content Report',
              body: body,
              labels: ['documentation', 'maintenance']
            });
Content Lifecycle
text
                 ┌─────────────┐
                 │   DRAFT     │  ← Created by contributor
                 └──────┬──────┘
                        │
                 ┌──────┴──────┐
                 │  REVIEW     │  ← Peer/SME review
                 └──────┬──────┘
                        │
                 ┌──────┴──────┐
                 │  PUBLISHED  │  ← Live in knowledge base
                 └──────┬──────┘
                        │
              ┌─────────┴─────────┐
              │                   │
       ┌──────┴──────┐    ┌──────┴──────┐
       │   UPDATED   │    │  ARCHIVED   │
       └──────┬──────┘    └─────────────┘
              │
       ┌──────┴──────┐
       │  REMOVED    │
       └─────────────┘

Cross-Referencing Strategy

markdown
## Cross-Referencing Best Practices

### Types of References

**Related guides**: Links to other how-to guides on related topics
> See also: [How to Configure Authentication](./authentication.md)

**Prerequisites**: Links to content users need before starting
> **Prerequisites**: [Installation Guide](./installation.md), [Account Setup](./account-setup.md)

**Deeper dives**: Links from overviews to detailed content
> For more detail, see [API Reference](../reference/api.md)

**Troubleshooting links**: Connect error messages to solutions
> If you see "Error 403: Forbidden", see [Troubleshooting Access Issues](./troubleshooting.md#403)

### Implementation

```markdown
## Resetting a User Password

> **Prerequisites**:
> - [Admin access setup](./admin-access.md)
> - User's email address or user ID

1. Log into the [Admin Dashboard](https://admin.example.com)
2. Navigate to **Users > Search**
3. Find the user and click **Reset Password**
4. The system sends a password reset email

> **Related**: [Bulk User Management](./bulk-user-management.md) |
> [Troubleshooting: Reset email not received](./troubleshooting.md#reset-email)

---

### Contribution Workflows

#### CONTRIBUTING.md for Knowledge Base

```markdown
# Contributing to the Knowledge Base

We welcome contributions from everyone. Here's how to add or improve content.

## Quick Start

1. Fork the knowledge base repository
2. Create a branch: `git checkout -b docs/my-new-guide`
3. Write your content in Markdown
4. Submit a Pull Request

## Content Standards

- **Title**: Clear, question-based or task-based title
- **Description**: One-line summary in frontmatter
- **Tags**: At least 2 tags (content type + topic)
- **Structure**: Prerequisites → Steps → Verification → Troubleshooting
- **Examples**: Include at least one runnable command or code snippet

## Templates

Use the template at `docs/_templates/guide.md` for new guides.

## Review Process

1. **Automated checks**: Markdown linting, link checking, spell check
2. **Peer review**: At least one team member reviews for accuracy
3. **Editorial review**: Content style and structure check (for major additions)
4. **Merge**: Squash and merge to main

## What Gets Accepted

| Type | Accepted? | Review Level |
|------|-----------|-------------|
| New guide | ✅ Yes | Full review |
| Typos/corrections | ✅ Yes | Quick review |
| Code example updates | ✅ Yes | SME review |
| Major rewrites | ✅ Yes | Full review |
| Duplicate content | ❌ No | — |
| Personal opinions | ❌ No | — |
| Outdated information | ⚠️ Requires update | Full review |
Contribution Metrics Dashboard
markdown
## Contributor Dashboard

| Month | Contributors | New Pages | Updates | Top Contributor |
|-------|-------------|-----------|---------|-----------------|
| Jan 2024 | 12 | 8 | 34 | @alice |
| Feb 2024 | 15 | 6 | 42 | @bob |
| Mar 2024 | 18 | 11 | 55 | @carol |

Analytics and Feedback

User Feedback Widget
markdown
<!-- Add to every page -->
## Was this page helpful?

- [✅ Yes] [❌ No]

<!-- If No -->
**What could we improve?**
[Free text field]

<!-- Footer -->
Last updated: {{ git_revision_date }}
[Edit this page]({{ edit_url }})
Analytics Tracking
javascript
// Docusaurus analytics integration
// docusaurus.config.js
module.exports = {
  themeConfig: {
    // Plausible (privacy-focused analytics)
    plausible: {
      domain: 'docs.myproject.com',
    },
    // Or Google Analytics
    googleAnalytics: {
      trackingID: 'UA-XXXXX-Y',
      anonymizeIP: true,
    },
  },
};
Key Metrics to Track
MetricWhat It Tells YouAction When Low
Search success rateAre people finding what they need?Improve content, tags, search config
Page views per articleWhich content is most/least used?Promote popular, fix or archive unpopular
Time on pageAre people reading or bouncing?Improve content quality, add examples
Feedback scoreAre users satisfied?Address negative feedback patterns
Return visitsIs the KB sticky?Add more cross-references, update content
Contribution rateIs the community engaged?Lower barriers, celebrate contributors

Knowledge Base Governance

Roles and Responsibilities
RoleResponsibilitiesWho
KB OwnerOverall strategy, roadmap, metricsPM or Tech Lead
Content StewardsContent quality, templates, style guideTech Writer
Subject Matter ExpertsTechnical accuracy, review, contributionEngineers
EditorsReview, merge, publishingSenior contributors
All ContributorsWrite, update, fix contentEveryone
Content Review Schedule
yaml
# Content review cadence by type
review_schedule:
  critical:
    - security procedures: monthly
    - incident response: monthly
    - authentication docs: quarterly
  standard:
    - guides and tutorials: quarterly
    - api reference: with each release
    - faq: semi-annually
  minor:
    - internal procedures: annually
    - project overviews: annually

Common Mistakes

  1. Building without a search strategy: A knowledge base with poor search is a graveyard. Invest in search from day one.

  2. No content owners: Pages without an owner go stale. Every page should have a clearly documented owner.

  3. Too many platforms: Don't split knowledge across Notion, Confluence, and a wiki. Pick one and migrate everything.

  4. Over-organizing: Creating too many categories and subcategories buries content. Flat structures with good tags outperform deep hierarchies.

  5. Perfectionism over publishing: Docs that are 80% complete and published are infinitely better than perfect docs that never ship.

  6. Not tagging content: A well-tagged page is discoverable. An untagged page is lost. Enforce tagging in your workflow.

  7. Ignoring stale content: Outdated documentation actively harms trust. Implement automated stale content detection.

  8. No contribution guidelines: If people don't know how to contribute, they won't. Make it trivially easy.

  9. No feedback loop: Without user feedback, you're writing in the dark. Add a "Was this helpful?" widget on every page.

  10. Treating the KB as a side project: Allocate dedicated time for KB maintenance, or it will decay into an abandoned graveyard.


Evaluation Rubric

Criterion1 - Graveyard2 - Collection3 - Organized4 - Curated5 - Living
PlatformNo platformBasic wikiPurpose-built KBOptimized KBIntegrated ecosystem
StructureNo structureBasic foldersTaxonomy definedInformation architectureAdaptive, user-validated IA
SearchNoneText searchTagged + text searchFaceted searchAI-assisted semantic search
Content FreshnessNever updatedUpdated reactivelyScheduled reviewsAutomated stale detectionContinuous improvement
ContributionNo processA few contributorsDefined workflowContributor programEmbedded in daily workflow
GovernanceNo ownershipUnclear ownershipDefined rolesEditorial boardMetrics-driven governance
FeedbackNoneOccasionalFeedback widgetAnalytics-drivenPredictive content recommendations

Frequently asked questions

What does the Knowledge Base AI skill do?

Knowledge Base Creation: Searchable wikis, internal documentation portals, knowledge management systems

Why use Knowledge Base on TypingMind?

Because you install it once and use it with any model. Knowledge Base 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 Knowledge Base in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/cosmicstack-labs/mercury-agent-skills/tree/main/categories/development/knowledge-base. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Knowledge Base?

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 Knowledge Base?

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

Is the Knowledge Base AI skill free?

Yes. It is published on GitHub by cosmicstack-labs under the MIT license. 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 👇