generate-llms
Generate llms.txt and llms-full.txt files for AI agent consumption following the llmstxt.org standard. Use when updating site content that should be reflected in the llms files, or when building/deploying the site.
Best use case
generate-llms is best used when you need a repeatable AI agent workflow instead of a one-off prompt.
Generate llms.txt and llms-full.txt files for AI agent consumption following the llmstxt.org standard. Use when updating site content that should be reflected in the llms files, or when building/deploying the site.
Teams using generate-llms should expect a more consistent output, faster repeated execution, less prompt rewriting.
When to use this skill
- You want a reusable workflow that can be run more than once with consistent structure.
When not to use this skill
- You only need a quick one-off answer and do not need a reusable workflow.
- You cannot install or maintain the underlying files, dependencies, or repository context.
Installation
Claude Code / Cursor / Codex
Manual Installation
- Download SKILL.md from GitHub
- Place it in
.claude/skills/generate-llms/SKILL.mdinside your project - Restart your AI agent — it will auto-discover the skill
How generate-llms Compares
| Feature / Agent | generate-llms | Standard Approach |
|---|---|---|
| Platform Support | Not specified | Limited / Varies |
| Context Awareness | High | Baseline |
| Installation Complexity | Unknown | N/A |
Frequently Asked Questions
What does this skill do?
Generate llms.txt and llms-full.txt files for AI agent consumption following the llmstxt.org standard. Use when updating site content that should be reflected in the llms files, or when building/deploying the site.
Where can I find the source code?
You can find the source code on GitHub using the link provided at the top of the page.
SKILL.md Source
# Generate LLMs Files
Generate `/llms.txt` and `/llms-full.txt` files that help AI agents understand and use the agentconfig.org site content.
## Overview
This skill generates machine-readable documentation files following the [llmstxt.org](https://llmstxt.org) standard:
- **`/llms.txt`** — A curated table of contents with links to detailed pages
- **`/llms-full.txt`** — Comprehensive markdown containing all site content
- **`/*.md`** — Page-specific markdown files (skills.md, agents.md, mcp.md, etc.)
## Page Registry
Pages are automatically discovered from the **page registry** at `site/src/data/pages.ts`:
```typescript
export const pages: readonly PageMeta[] = [
{
slug: 'skills',
title: 'Skills Tutorial',
description: 'How to create agent skills...',
mdFile: 'skills.md',
partNumber: 3,
},
// ... more pages
]
```
When creating a new page, add an entry to the registry to include it in the llms files.
## When to Use
Use this skill when:
- You've updated content in the data files (`site/src/data/*.ts`)
- You've added a new page (remember to add it to the page registry!)
- Before deploying or releasing a new version of the site
- When an agent needs to understand what content should be in llms files
## Data Sources
The llms files are generated from these TypeScript data files:
| File | Content |
|------|---------|
| `site/src/data/pages.ts` | **Page registry** - lists all pages for llms generation |
| `site/src/data/primitives.ts` | 10 AI primitives with descriptions, use cases, and provider implementations |
| `site/src/data/comparison.ts` | Provider comparison matrix (Copilot vs Claude) |
| `site/src/data/skillsTutorial.ts` | Skills tutorial sections and concepts |
| `site/src/data/skillExamples.ts` | 5 example skills with full code |
| `site/src/data/agentsTutorial.ts` | Agent definitions tutorial with code samples |
| `site/src/data/mcpTutorial.ts` | MCP tutorial sections and code samples |
## File Structure
```
site/public/
├── llms.txt # Table of contents (llmstxt.org format)
├── llms-full.txt # Complete site content in one file
├── skills.md # Skills page content (rich markdown)
├── agents.md # Agents page content (rich markdown)
└── mcp.md # MCP page content (rich markdown)
```
## Generation Process
### Step 1: Generate `/llms.txt` (Table of Contents)
The homepage llms.txt follows the strict llmstxt.org format:
- H1 with project name
- Blockquote with summary
- Optional paragraph (no headings)
- H2 sections with link lists only
```markdown
# agentconfig.org
> A reference site for configuring AI coding assistants. Covers 10 AI primitives,
> provider comparison (GitHub Copilot vs Claude Code), and tutorials for skills
> and agent definitions.
## Pages
- [Homepage](https://agentconfig.org/): AI primitives reference, file tree, provider comparison
- [Skills Tutorial](https://agentconfig.org/skills): How to create agent skills following agentskills.io
- [Agents Tutorial](https://agentconfig.org/agents): Agent definition files for Copilot and Claude
## Docs
- [Full site content](/llms-full.txt): Complete content for deep context
- [Skills page content](/skills.md): Skills tutorial with 5 example skills
- [Agents page content](/agents.md): Agents tutorial with code samples
## Optional
- [agentskills.io specification](https://agentskills.io/specification): The skills format
- [AGENTS.md specification](https://agents.md): Open format for coding agents
```
### Step 2: Generate `/llms-full.txt` (Complete Content)
Run the generation script:
```bash
bun .github/skills/generate-llms/scripts/generate-llms-full.ts
```
This reads all data files and generates comprehensive markdown including:
1. **Site overview** with purpose and structure
2. **All 10 AI primitives** with full details:
- Description
- What it is
- When to use it
- What it prevents
- What to combine it with
- Provider implementations (Copilot and Claude)
3. **Provider comparison table** showing support levels
4. **Config file locations** for both providers
5. **Skills tutorial** with all concepts and 5 example skills
6. **Agents tutorial** with all sections and code samples
### Step 3: Generate Page-Specific `.md` Files
**`/skills.md`:**
- Tutorial sections (Understanding the Spec, Progressive Disclosure, etc.)
- 5 example skills with full SKILL.md content
- Key takeaways for each example
**`/agents.md`:**
- 9 tutorial sections from beginner to advanced
- Code samples for AGENTS.md, CLAUDE.md, copilot-instructions.md
- Path-scoped rules, agent personas, monorepo strategies
## Manual Generation
If the script isn't available, generate manually by:
1. Read each data file in `site/src/data/`
2. Extract the exported arrays/objects
3. Format as markdown following the templates below
### Template: Primitive Entry
```markdown
### {name}
{description}
**What it is:** {whatItIs}
**Use when:**
{useWhen as bullet list}
**Prevents:** {prevents}
**Combine with:** {combineWith as comma-separated list}
**Provider Implementations:**
| Provider | Implementation | Location | Support |
|----------|---------------|----------|---------|
| Copilot | {implementation} | {location} | {support} |
| Claude | {implementation} | {location} | {support} |
```
### Template: Skill Example
```markdown
### {displayName}
**Complexity:** {complexity}
**Demonstrates:** {demonstrates}
{description}
**Files:**
\`\`\`{language}
// {path}
{content}
\`\`\`
**Key Takeaways:**
{keyTakeaways as bullet list}
```
## Validation
After generating, verify:
1. **Links work** — All internal links resolve
2. **Content is current** — Matches what's shown on the site
3. **Format is correct** — Valid markdown, proper headings
4. **Size is reasonable** — llms-full.txt should be comprehensive but not bloated
## Example Prompts
### Generate all llms files
```
Generate all the llms.txt files for agentconfig.org
```
### Update after content change
```
I updated the primitives data, regenerate the llms files
```
### Check what should be in llms-full.txt
```
What content should be included in llms-full.txt?
```Related Skills
building-with-llms
Help users build effective AI applications. Use when someone is building with LLMs, writing prompts, designing AI features, implementing RAG, creating agents, running evals, or trying to improve AI output quality.
nano-image-generate
Generate images using Nano Banana (Flash) or Nano Banana Pro. Use 'flash' for speed/efficiency and 'pro' for high quality, text rendering, and complex prompt adherence. Triggers include 'generate image', 'create logo', 'fast image', 'high quality image'.
Generate educational content with fixed sections
Generate educational content about any topic following a user-specified structure: include a set number of learning tips/tricks or activities, fun facts, Q&A pairs, jokes, and optionally extend with additional lines. Always list supplies for activities when requested.
fal-generate
Generate images and videos using fal.ai AI models
generate-component-story
Create story examples for components. Use when writing stories, creating examples, or demonstrating component usage.
agentuity-cli-project-auth-generate
Generate SQL schema for Agentuity Auth tables. Use for managing authentication credentials
working-with-llms
Mandatory workflow for creating LLM-facing content. Follow the 4-step process (objective → draft → verify → adjust) before writing any prompt, skill, tool description, or system instruction. Triggers on requests to create or revise skills, prompts, agent workflows, or any content that will be sent to an LLM repeatedly.
serving-llms-vllm
Serves LLMs with high throughput using vLLM's PagedAttention and continuous batching. Use when deploying production LLM APIs, optimizing inference latency/throughput, or serving models with limited GPU memory. Supports OpenAI-compatible endpoints, quantization (GPTQ/AWQ/FP8), and tensor parallelism.
bgo
Automates the complete Blender build-go workflow, from building and packaging your extension/add-on to removing old versions, installing, enabling, and launching Blender for quick testing and iteration.
humanizer-ko
Detects and corrects Korean AI writing patterns to transform text into natural human writing. Based on scientific linguistic research (KatFishNet paper with 94.88% AUC accuracy). Analyzes 19 patterns including comma overuse, spacing rigidity, POS diversity, AI vocabulary overuse, and structural monotony. Use when humanizing Korean text from ChatGPT/Claude/Gemini or removing AI traces from Korean LLM output.
huggingface-accelerate
Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision (FP16/BF16/FP8). Interactive config, single launch command. HuggingFace ecosystem standard.
hr-pro
Professional, ethical HR partner for hiring, onboarding/offboarding, PTO and leave, performance, compliant policies, and employee relations. Ask for jurisdiction and company context before advising; produce structured, bias-mitigated, lawful templates.