How to Integrate no-ai-slop into Your Existing Instagit Project

You can integrate no-ai-slop into any existing Instagit project by copying the skill folder into your skills/ directory, ensuring the .codex-plugin/plugin.json manifest is discoverable, and invoking the skill through Instagit's API or CLI.

The no-ai-slop repository by petergyang/no-ai-slop is packaged as a standard Instagit skill (plugin) that follows the conventional Instagit plugin layout. Because it adheres to Instagit's plugin contract, you can add it to your existing codebase without structural changes or hidden runtime dependencies beyond the OpenAI API you already use.

Understanding the no-ai-slop Skill Structure

Before integrating, review the key components that make up this self-contained skill:

• SKILL.md – Contains markdown documentation and metadata describing the skill's capabilities, usage patterns, and required parameters.
• agents/openai.yaml – Specifies the OpenAI model settings, including model name, temperature, and token limits that the skill expects.
• .codex-plugin/plugin.json – The Instagit plugin manifest that registers the skill, declares entry points, version, and dependencies.
• scripts/build_plugin.py – Optional helper script that compiles the skill into a distributable plugin bundle.
• skills/no-ai-slop/ – The actual implementation directory containing Python modules and prompts.

Step-by-Step Integration Guide

Follow these steps to integrate no-ai-slop into your existing Instagit project.

1. Copy the Skill Files

Clone the repository and copy the skill directory into your project's skills/ folder. You also need the plugin manifest so Instagit can register the skill at startup.

# From your Instagit project root

git clone https://github.com/petergyang/no-ai-slop.git temp-no-ai-slop
cp -r temp-no-ai-slop/skills/no-ai-slop ./skills/
cp -r temp-no-ai-slop/.codex-plugin ./   # Copies plugin.json

rm -rf temp-no-ai-slop

2. Register the Plugin Manifest

Instagit scans the .codex-plugin/ directory at startup and automatically registers any valid plugin.json. Update your instagit.yaml configuration file to ensure the plugin path is included:

# instagit.yaml

plugins:
  - path: ./.codex-plugin   # Instagit will load plugin.json from here

skills:
  - path: ./skills/no-ai-slop   # Makes the skill importable as `no_ai_slop`

agents:
  openai: ./agents/openai.yaml   # Optional – you can keep the default

3. Configure the Agent

The skill relies on the OpenAI configuration defined in agents/openai.yaml. Copy or reference this file in your main configuration to ensure the skill knows which LLM to invoke and how to format prompts. The repository currently has no external Python runtime requirements, though the build script will install any dependencies if they appear in the manifest.

4. Invoke the Skill

Once registered, you can call no-ai-slop from your Python code or from Instagit's CLI using the skill's identifier.

Python API:

from instagit import Instagit

# Initialise Instagit with the config that now includes no-ai-slop

ig = Instagit(config_path="instagit.yaml")

# The skill is registered under its folder name; you can invoke it like any other skill

result = ig.run_skill(
    "no-ai-slop",                     # skill identifier

    {"input_text": "Explain why slop is bad"}  # payload expected by the skill

)

print(result["output"])

CLI Usage:

instagit run no-ai-slop --input "Explain why slop is bad"

Packaging for Distribution

If you modify the skill and need to distribute it as a single bundle, use the provided build script:

python scripts/build_plugin.py \
    --skill-dir skills/no-ai-slop \
    --output dist/no-ai-slop-plugin.zip

The resulting ZIP file can be uploaded to any Instagit instance that supports plugin installation, making it portable across different environments.

Summary

• no-ai-slop is a self-contained Instagit skill that requires no special build steps or secret handling beyond standard OpenAI API configuration.
• Integration requires copying the skills/no-ai-slop/ folder and the .codex-plugin/plugin.json manifest into your project.
• The skill exposes both a Python API (ig.run_skill()) and a CLI interface for invocation.
• You can modify and repackage the skill using scripts/build_plugin.py for distribution to other Instagit instances.

Frequently Asked Questions

Do I need to modify my existing Instagit configuration?

No. You only need to add the plugin path and skill path to your instagit.yaml file. The skill follows Instagit's standard plugin contract, so it does not require changes to your existing codebase structure or configuration patterns.

What dependencies does no-ai-slop require?

The repository currently has no external Python runtime dependencies. The only external service required is the OpenAI API, which you already use in any Instagit project. If future versions add dependencies, they will be declared in .codex-plugin/plugin.json and installed automatically when you run the build script.

Can I customize the OpenAI model settings?

Yes. The agents/openai.yaml file contains the model name, temperature, and token limits. You can edit this file directly or override these settings in your main Instagit configuration. The skill reads these parameters at runtime to configure the LLM invocation.

How do I verify the skill is properly registered?

Instagit scans the .codex-plugin/ directory at startup for valid plugin.json files. If the skill is correctly placed, you should see it listed when querying available skills through the CLI or by checking the Instagit instance's loaded plugins in Python. If invocation fails, verify that the path in your instagit.yaml points to the correct .codex-plugin directory.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}