Cognotik

The Cognotik Build Integration Demo showcases how to integrate Cognotik as a build and automation tool using GitHub Actions. This repository provides working examples of AI-powered code review, automated bug fixing, and intelligent issue resolution integrated directly into your CI/CD pipeline.

Table of Contents

• Features
• Quick Start
• Usage
• What is Cognotik?
• GitHub Actions Workflows
  • Auto-Fixing Build Validator
  • Agentic Issue Handler
• Configuration
  • Code Reviewer
  • Code Fixer
  • Code Implementer
• API Providers
• Contributing
• License

Features

• 🔄 Ready-to-Use GitHub Actions: Pre-configured workflows for immediate integration
• 🔧 Automated Build Fixing: Automatically fix failing builds in pull requests
• 🎫 Intelligent Issue Resolution: AI agents that create PRs to resolve GitHub issues
• 🤖 Multi-Model Support: Works with OpenAI, Google Gemini, Anthropic Claude, and Groq
• 📝 Code Review Automation: Review code against best practices and standards
• 🏗️ Code Generation: Generate implementations from natural language descriptions
• 📚 Documentation Processing: Automatically update code and docs based on markdown specifications

Quick Start

1. Fork or Clone This Repository

git clone https://github.com/SimiaCryptus/CognotikDemo.git
cd CognotikDemo

2. Configure Repository Secrets

Add the following secrets to your GitHub repository (Settings → Secrets and variables → Actions):

• OPENAI_API_KEY - Your OpenAI API key
• GOOGLE_API_KEY - Your Google Gemini API key (recommended)

3. Enable GitHub Actions

The workflows in .github/workflows/ will automatically:

• Fix failing builds in pull requests
• Create PRs to resolve issues labeled with agent-help

4. Local Development (Optional)

export GOOGLE_API_KEY="your-google-api-key"

./gradlew codeReview -PreviewPrompt="Review and improve code quality in file (%s)"

What is Cognotik?

Cognotik is a powerful AI-powered code automation framework that enables intelligent code review, generation, and modification using large language models (LLMs). This demo project shows how to leverage Cognotik's capabilities within GitHub Actions workflows.

Key Cognotik Components Used

• Code Reviewer: Analyzes source files against standards and suggests improvements
• Code Fixer: Parses build logs and automatically fixes compilation/test errors
• Code Implementer: Generates new code from natural language descriptions

GitHub Actions Workflows

This demo includes two pre-configured GitHub Actions workflows that showcase Cognotik's automation capabilities.

Auto-Fixing Build Validator

File: .github/workflows/autofix_build.yml

Automatically detects and fixes failing builds in pull requests.

How It Works

1. Triggered on pull requests to main
2. Runs the test suite and captures output to build.log
3. If tests fail, invokes Cognotik's Code Fixer agent
4. Agent analyzes the build log and fixes implementation files
5. Verifies the fix by re-running tests
6. Commits the fixes automatically

name: Auto-Fixing Build Validator

on:
  pull_request:
    branches: [ main ]


jobs:
  build-and-fix:
    runs-on: ubuntu-latest
    steps:


      - name: Run Tests
        id: run_tests
        run: ./gradlew test 2>&1 | tee build.log
        continue-on-error: true

      - name: Agentic Auto-Fix
        if: steps.run_tests.outcome == 'failure'
        run: |

          ./gradlew codeFixer \
            -PfixPrompt="Fix the build errors" \
            -PfixLog="build.log"

Agentic Issue Handler

File: .github/workflows/issue_handler.yml

Automatically creates pull requests to resolve GitHub issues labeled with agent-help.

How It Works

1. Triggered when an issue is opened or labeled with agent-help
2. Analyzes the issue title and description
3. Invokes Cognotik's Code Implementer agent
4. Agent reviews relevant source files and generates fixes
5. Creates a pull request with the proposed changes
6. Comments on the issue if no changes were needed

Usage

1. Create a new issue describing the bug or feature
2. Add the agent-help label
3. Wait for the agent to analyze and create a PR

Configuration

Required Repository Secrets

Secret	Description	Required
GOOGLE_API_KEY	Google Gemini API key	Yes*
OPENAI_API_KEY	OpenAI API key	Yes*

*At least one API key is required. Google Gemini is recommended for best results.

Optional Secrets

Secret	Description
ANTHROPIC_API_KEY	Anthropic Claude API key
GROQ_API_KEY	Groq API key

Usage

Gradle Tasks

This demo provides three main Gradle tasks:

Code Reviewer

./gradlew codeReview \
  -PreviewPrompt="Update implementation file (%s) according to best practices" \
  -PreviewSrc="src/main/java" \
  -PreviewThreads=4

Code Fixer

./gradlew codeFixer \
  -PfixPrompt="Fix the build errors reported in build.log" \
  -PfixLog="build.log"

Code Implementer

./gradlew codeImplementer \
  -PimplPrompt="Create a REST API endpoint for user management" \
  -PimplHeadless=true

Documentation Processor

./gradlew docProcessor \
  -PoverwriteMode="PatchToUpdate" \
  -ProotDir="." \
  -Pthreads=4

Documentation Processing

The DocProcessor enables bidirectional synchronization between documentation and code using markdown frontmatter specifications.

How It Works

Create markdown files with YAML frontmatter that specify relationships to source files:

Specifying Target Files

Use the specifies key to indicate which files should be updated based on the documentation:

---
specifies:
  - "../src/main/java/com/example/*.java"
  - "../src/main/kotlin/**/*.kt"
---
# API Design Guidelines
All service classes should follow these patterns...

Documenting Source Files

Use the documents key to indicate which source files should inform the documentation:

---
documents:
  - "../src/main/java/com/example/UserService.java"
  - "../src/main/java/com/example/UserRepository.java"
---
# User Service API Documentation
This documentation is automatically updated based on the source files...

File Transformations

Use the transforms key to specify source-to-destination file mappings with regex patterns:

---
transforms:
  - "src/main/java/(.+)\\.java -> src/test/java/$1Test.java"
  - "src/main/kotlin/(.+)\\.kt -> docs/api/$1.md"
---
# Test Generation Template
Generate test files based on the source implementation...

Frontmatter Options

Key	Description	Example
specifies	Glob patterns for files to update based on this doc	"../src/**/*.java"
documents	Glob patterns for source files that inform this doc	"../src/main/**/*.kt"
transforms	Regex patterns with backreferences for file transformations	"src/(.+)\\.java -> test/$1Test.java"

Overwrite Modes

Mode	Description
SkipExisting	Skip files that already exist
OverwriteExisting	Completely replace existing files
OverwriteToUpdate	Replace files only if source is newer than target
PatchExisting	Apply incremental patches to existing files
PatchToUpdate	Apply incremental patches only if source is newer than target (default)

API Providers

Cognotik supports multiple AI providers. Configure your preferred provider:

Provider	Model Example	Environment Variable
Google	Gemini 3.0	GOOGLE_API_KEY
OpenAI	GPT-4o	OPENAI_API_KEY
Anthropic	Claude 4.5	ANTHROPIC_API_KEY
Groq	Llama 3 70B	GROQ_API_KEY

Contributing

We welcome contributions! Please see our Contributing Guide for details.

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

---

Support

• 📖 Cognotik Documentation
• 🐛 Issue Tracker
• 💬 Discussions

---

Made with ❤️ by SimiaCryptus | Powered by Cognotik

• 📚 Documentation Processing: Automatically update code and docs based on markdown specifications