synthesis-llm-setup v2.0.0A

Immutable. This exact content never changes and is served at /api/v1/blob/2926d62a4aa92794.

---
name: synthesis-llm-setup
description: "Guide for setting up and configuring LLM projects across platforms. Use when setting up a Claude Project, creating a ChatGPT GPT, configuring a Gemini Gem, setting up an AI assistant with custom instructions and knowledge, or configuring any LLM platform with AI Knowledge content."
license: "CC0-1.0"
depends_on: []
metadata:
  author: "Rajiv Pant"
  version: "2.0.0"
  format: v5
  source_repo: "github.com/synthesisengineering/synthesis-skills"
  source_type: "public"
---

# LLM Setup

Configure Claude Projects, ChatGPT GPTs, Gemini Gems, and other LLM platforms using compiled AI Knowledge content from the ragbot system.

## Binding rules

1. **Edit `source/` files, never the generated output.** CI/CD rebuilds `all-knowledge.md` on every push to `source/`, so a hand edit to it is overwritten.
2. **Take instructions from `compiled/{project}/instructions/`:** the file for that platform (`claude.md`, `chatgpt.md`, `gemini.md`), or the closest match for any other LLM.
3. **Recompile instructions only when they change** (`ragbot compile --project {name}`); knowledge concatenation needs no manual step.
4. **Deliver knowledge per platform:** GitHub sync for Claude; upload `all-knowledge.md` to ChatGPT and Gemini, and upload it again after source changes.
5. **Compile where the audience matches.** A compilation includes only repos the compiling user can access, which is what keeps personal content out of team and client projects.
6. **Verify each project** with a representative query before calling it set up.

## Contents

- [references/platform-setup.md](references/platform-setup.md): click-by-click setup for Claude Projects, ChatGPT GPTs, Gemini Gems and other LLMs. Read when creating or configuring a project on one of them.
- [references/compilation-and-inheritance.md](references/compilation-and-inheritance.md): how compilation and inheritance decide what content each project gets, the privacy model, and the `ragbot compile` options. Read before compiling, or when choosing which repo to compile in.
- [references/troubleshooting.md](references/troubleshooting.md): fixes for instructions too long, knowledge not used, inheritance not working, and content from the wrong repo. Read when a project misbehaves.
- [references/coverage-map.md](references/coverage-map.md): where each part of the 1.0.0 text now lives (ruling D8).
- Configuration, Architecture overview, Output structure: below. The settings to adjust and what the system produces.
- Step-by-step workflow, Updating projects: below. The procedure for a new project and for keeping one current.

## Configuration

These values are user-specific. Update them for your environment.

| Setting | Value | Description |
|---------|-------|-------------|
| `ai_knowledge_repo` | `ai-knowledge-{name}/` | Your ai-knowledge repository root |
| `compiled_instructions_path` | `compiled/{project}/instructions/` | Path to LLM-specific compiled instructions |
| `knowledge_file` | `all-knowledge.md` | Concatenated knowledge file (auto-generated by CI/CD) |
| `source_dir` | `source/` | Directory to edit directly (knowledge concatenation is automatic) |

## Architecture overview

The AI Knowledge system produces platform-specific content for each LLM's project/custom instruction system.

| Operation | Where | When |
|-----------|-------|------|
| Knowledge concatenation (`all-knowledge.md`) | CI/CD (GitHub Actions) | Every push to `source/` |
| Instruction compilation | Local (`ragbot compile`) | When instructions change (rare) |
| RAG indexing | Local (`ragbot index`) | When content changes and RAG is needed |

## Output structure

```
ai-knowledge-{name}/
├── compiled/
│   └── {project}/
│       └── instructions/         # LLM-specific custom instructions
│           ├── claude.md
│           ├── chatgpt.md
│           └── gemini.md
└── all-knowledge.md              # Concatenated knowledge (auto-generated by CI/CD)
```

**Key principle:** Edit `source/` files directly. Knowledge concatenation is automatic.

**Knowledge delivery strategy:**
- Claude: GitHub sync the repo (Claude reads `all-knowledge.md` directly)
- ChatGPT: Upload `all-knowledge.md`
- Gemini: Upload `all-knowledge.md` (works within the 10-file Gem limit)

## Step-by-step workflow

1. **Compile instructions** (only if instructions changed):
   ```bash
   ragbot compile --project {name} --no-llm
   ```

2. **For each project** (e.g., client-a):
   - **Claude**: Create project, copy `compiled/client-a/instructions/claude.md` to custom instructions, GitHub sync the repo
   - **ChatGPT**: Create GPT, copy `compiled/client-a/instructions/chatgpt.md` to instructions, upload `all-knowledge.md`
   - **Gemini**: Create Gem, copy `compiled/client-a/instructions/gemini.md` to instructions, upload `all-knowledge.md`

3. **Verify** by testing each project with a representative query.

## Updating projects

### When to update

**Knowledge (all-knowledge.md):** Updates automatically via CI/CD on every push to `source/`. No manual step.

**Instructions:** Recompile only when instructions change (rare):
```bash
ragbot compile --project {name} --force
```

**LLM sync:**
- Claude: GitHub sync auto-updates
- ChatGPT/Gemini: Re-upload `all-knowledge.md` after source changes