Hefesto AI - Intelligent Code Quality Analysis and Technical Debt Detection
Project description
๐ฅ Hefesto - AI-Powered Code Quality Guardian
Multi-language AI code quality validation powered by Gemini that works seamlessly with AI coding assistants.
Hefesto is the AI-powered code quality guardian that validates your Python, TypeScript, and JavaScript code before it hits production. It caught 3 critical bugs in its own v4.0.1 release through self-validation (dogfooding). Now it's protecting codebases worldwide.
๐ฏ What is Hefesto?
Hefesto analyzes your code using AI and catches issues that traditional linters miss:
- Multi-Language Support: Python, TypeScript, JavaScript (Java & Go coming soon)
- AI-Powered Analysis: Uses Google Gemini for semantic code understanding
- ML Enhancement: Detects code smells, duplications, and anti-patterns
- Security Scanning: Finds hardcoded secrets, SQL injections, command injections
- Pre-Push Hooks: Validates code automatically before every commit
- REST API: Integrate into any workflow
- BigQuery Analytics: Track code quality over time
๐ OMEGA Guardian
OMEGA Guardian adds production monitoring on top of Hefesto PRO:
- IRIS Agent: Real-time production monitoring
- Auto-Correlation: Links code issues to production incidents
- Real-Time Alerts: Get notified when code causes production failures
- Unlimited Everything: Repos, LOC, users - no limits
๐ The Dogfooding Story
We used Hefesto to validate Hefesto itself.
Before publishing v4.0.1 to PyPI, we ran OMEGA Guardian's self-validation on its own codebase. Here's what it caught:
๐ Critical Bugs Found:
-
Hardcoded Password in test fixtures
- Severity: CRITICAL
- Location:
tests/fixtures/auth.py - Could leak credentials to GitHub
-
Dangerous
exec()Call without validation- Severity: HIGH
- Location:
utils/dynamic_loader.py - Remote code execution vulnerability
-
155 Other Issues including:
- 23 code smells
- 12 security warnings
- 47 complexity violations
- 73 best practice violations
โ Result:
We fixed everything before shipping. v4.0.1 went to production clean.
This is meta-validation at its finest: AI validating AI code.
๐ฐ Pricing - Launch Special
๐ Lock in Launch Pricing Forever
First 100 customers get launch pricing locked permanently. Sign up now and your rate never increases.
| Feature | FREE | PRO | OMEGA Guardian |
|---|---|---|---|
| Price | $0 | $8/mo | $19/mo |
| Future Price | $0 | $25/mo | $35/mo |
| You Save | - | $204/year | $192/year |
| Basic Analysis | โ | โ | โ |
| Pre-push Hooks | โ | โ | โ |
| CLI Commands | โ | โ | โ |
| AI/ML Enhancement | โ | โ | โ |
| REST API | โ | โ | โ |
| BigQuery Integration | โ | โ | โ |
| IRIS Monitoring | โ | โ | โ |
| Production Correlation | โ | โ | โ |
| Real-time Alerts | โ | โ | โ |
| Repos/LOC/Users | Limited | Unlimited | Unlimited |
๐ 14-Day Free Trial
Both PRO and OMEGA Guardian include 14 days free trial. No credit card required upfront.
Get Started Now
๐ Hefesto PRO - $8/month
AI-powered code quality with ML enhancement
โ SemanticAnalyzer (AI/ML) โ REST API (8 endpoints) โ BigQuery integration โ Unlimited analysis โ Priority support โ 14-day free trial
๐ OMEGA Guardian - $19/month
Everything in PRO + Production Monitoring
โ Everything in Hefesto PRO โ IRIS Agent monitoring โ Auto-correlation engine โ Real-time alerts โ Production incident tracking โ Unlimited everything โ 14-day free trial
๐ Quick Start
Installation
# FREE tier
pip install hefesto-ai
# PRO tier ($8/month)
pip install hefesto-ai[pro]
# OMEGA Guardian ($19/month)
pip install hefesto-ai[omega]
Basic Usage
# Analyze a single file
hefesto analyze main.py
# Analyze entire directory
hefesto analyze .
# With severity filter
hefesto analyze . --severity HIGH
# JSON output
hefesto analyze . --format json
Activate PRO/OMEGA Features
# Set license key (from Stripe after purchase)
export HEFESTO_LICENSE_KEY="your-license-key-here"
# Verify activation
hefesto status
# Should show:
# License: PRO โ
(or OMEGA โ
)
# ML Enhancement: โ
Enabled
Pre-Push Hook (Automatic Validation)
# Install git hook
hefesto install-hook
# Now every git push validates your code automatically
git add .
git commit -m "Add new feature"
git push # โ Hefesto validates before pushing
๐ฏ Features
FREE Tier
- โ Static Analysis: Complexity, code smells, best practices
- โ Security Scanning: Hardcoded secrets, injections (basic)
- โ CLI Commands: Analyze, status, install hooks
- โ Pre-Push Hooks: Automatic validation on git push
- โ Multi-language: Python, JavaScript, TypeScript, Go, Rust
PRO Tier ($8/month)
Everything in FREE, plus:
- โ SemanticAnalyzer: AI-powered ML code understanding
- โ Deep Security Scanning: Advanced vulnerability detection
- โ REST API: 8 endpoints for CI/CD integration
- โ BigQuery Integration: Historical code quality analytics
- โ Duplicate Detection: Find copy-pasted code
- โ Anti-Pattern Detection: Identify design flaws
- โ Priority Support: Email support with 24h response
OMEGA Guardian ($19/month)
Everything in PRO, plus:
- โ IRIS Agent: Production monitoring and alerting
- โ HefestoEnricher: Auto-correlate code issues โ production failures
- โ Real-Time Alerts: Pub/Sub notifications when code causes incidents
- โ BigQuery Analytics: Track correlations over time
- โ Production Dashboard: Visualize code quality โ production health
- โ Unlimited Everything: Repos, LOC, users, analysis
- โ Priority Slack Support: Direct Slack channel access
๐ REST API
Hefesto PRO includes a REST API for CI/CD integration:
# Start server
hefesto serve --port 8000
# Analyze code via API
curl -X POST http://localhost:8000/analyze \
-H "Content-Type: application/json" \
-d '{"code": "def test(): pass", "severity": "MEDIUM"}'
API Endpoints
POST /analyze- Analyze codeGET /health- Health checkPOST /batch- Batch analysisGET /metrics- Quality metricsGET /history- Analysis historyPOST /webhook- GitHub webhook integrationGET /stats- StatisticsPOST /validate- Validate without storing
๐ง Configuration
License Key
Set via environment variable:
export HEFESTO_LICENSE_KEY="your-key-here"
Or create .hefesto.env:
HEFESTO_LICENSE_KEY=your-key-here
HEFESTO_SEVERITY=MEDIUM
HEFESTO_OUTPUT=json
Custom Rules
Create .hefesto.yaml:
severity: HIGH
exclude:
- tests/
- node_modules/
- .venv/
rules:
complexity:
max_cyclomatic: 10
max_cognitive: 15
security:
check_secrets: true
check_injections: true
ml:
enabled: true
threshold: 0.7
๐๏ธ OMEGA Guardian Setup
OMEGA Guardian requires Docker for the IRIS Agent:
1. Install OMEGA Guardian
pip install hefesto-ai[omega]
export HEFESTO_LICENSE_KEY="your-omega-key"
2. Configure IRIS Agent
Create iris_config.yaml:
project_id: your-gcp-project
dataset: omega_production
pubsub_topic: hefesto-alerts
alert_rules:
- name: error_rate_spike
query: |
SELECT COUNT(*) as error_count
FROM `production.logs`
WHERE severity = 'ERROR'
AND timestamp > TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 5 MINUTE)
threshold: 10
- name: latency_increase
query: |
SELECT AVG(latency_ms) as avg_latency
FROM `production.metrics`
WHERE timestamp > TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 5 MINUTE)
threshold: 1000
3. Run IRIS Agent
# Via Docker (coming soon)
docker run -v ./iris_config.yaml:/config.yaml \
narapa/iris-agent:latest
# Or via Python
python -m hefesto.omega.iris_agent --config iris_config.yaml
4. Verify Correlation
# Check that IRIS is correlating issues
hefesto omega status
# Should show:
# IRIS Agent: โ
Running
# Correlations: 3 active
# Last Alert: 2 minutes ago
๐งช Testing & CI/CD
GitHub Actions
name: Hefesto Validation
on: [push, pull_request]
jobs:
validate:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Install Hefesto
run: pip install hefesto-ai[pro]
- name: Run Analysis
env:
HEFESTO_LICENSE_KEY: ${{ secrets.HEFESTO_LICENSE_KEY }}
run: hefesto analyze . --severity HIGH --format json
GitLab CI
hefesto:
stage: test
script:
- pip install hefesto-ai[pro]
- export HEFESTO_LICENSE_KEY=$HEFESTO_LICENSE_KEY
- hefesto analyze . --severity HIGH
Pre-Commit Hook
# .pre-commit-config.yaml
repos:
- repo: local
hooks:
- id: hefesto
name: Hefesto Analysis
entry: hefesto analyze
language: system
pass_filenames: false
๐ Use Cases
1. AI Coding Assistant Validation
Use Hefesto to validate code generated by Claude, Cursor, or GitHub Copilot:
# After AI generates code
hefesto analyze generated_code.py --severity MEDIUM
# Before committing AI-generated code
git add .
git commit -m "AI-generated feature"
git push # โ Hefesto validates automatically
2. Production Monitoring (OMEGA)
Correlate code quality issues with production failures:
# IRIS detects production error spike
# HefestoEnricher correlates to recent code changes
# Alert sent: "High complexity function causing 500 errors"
3. Team Code Reviews
Run Hefesto before code review to catch obvious issues:
# Before opening PR
hefesto analyze feature_branch/ --format json > review.json
# Share review.json with team
4. Technical Debt Tracking
Track code quality over time with BigQuery:
-- Query code quality trends
SELECT
DATE(analyzed_at) as date,
AVG(complexity_score) as avg_complexity,
COUNT(*) as issues_found
FROM hefesto_analytics.analyses
WHERE project = 'my-app'
GROUP BY date
ORDER BY date DESC
๐ก๏ธ Security
Hefesto helps find security vulnerabilities:
What Hefesto Catches
- โ Hardcoded Secrets: API keys, passwords, tokens
- โ SQL Injection: Unsafe query construction
- โ Command Injection: Unsafe shell command execution
- โ Path Traversal: Unsafe file access
- โ Unsafe Deserialization: pickle, yaml.unsafe_load
- โ XSS Vulnerabilities: Unsafe HTML rendering
- โ SSRF Attempts: Unsafe URL requests
Example
# Hefesto catches this:
password = "admin123" # โ Hardcoded secret
os.system(f"rm {user_input}") # โ Command injection
query = f"SELECT * FROM users WHERE id={user_id}" # โ SQL injection
# Hefesto suggests:
password = os.getenv("PASSWORD") # โ
subprocess.run(["rm", user_input], check=True) # โ
cursor.execute("SELECT * FROM users WHERE id=?", (user_id,)) # โ
๐ Documentation
- Installation: Installation Guide
- Configuration: Configuration Guide
- API Reference: API Docs
- OMEGA Guardian: OMEGA Setup
- Examples: Examples Directory
๐ค Contributing
We welcome contributions! See CONTRIBUTING.md for guidelines.
Development Setup
# Clone repo
git clone https://github.com/artvepa80/Agents-Hefesto.git
cd Agents-Hefesto
# Install in development mode
pip install -e ".[dev]"
# Run tests
pytest tests/
# Run Hefesto on itself (dogfooding)
hefesto analyze . --severity MEDIUM
๐ Changelog
See CHANGELOG.md for version history.
Recent Releases
- v4.2.1 (2025-10-31): Critical tier hierarchy bugfix
- v4.2.0 (2025-10-31): OMEGA Guardian release
- v4.1.0 (2025-10-31): Unified package architecture
- v4.0.1 (2025-10-30): Production stability fixes
โ FAQ
Do I need a license for the FREE tier?
No, FREE tier works without any license key.
How do I upgrade from FREE to PRO?
- Purchase PRO: Get PRO
- Set license key:
export HEFESTO_LICENSE_KEY="your-key" - Features unlock automatically
Can I try PRO/OMEGA before buying?
Yes! Both include 14-day free trials. No credit card required upfront.
What happens after 100 launch customers?
Pricing increases to $25/mo (PRO) and $35/mo (OMEGA) for new customers. Early customers keep their launch pricing forever.
Is my code sent to external servers?
- FREE/PRO: Analysis runs locally, no code sent externally
- OMEGA: Only metadata sent to BigQuery for correlation
- Your actual code never leaves your infrastructure
What if I have issues?
- Email: support@narapallc.com
- GitHub Issues: Open an issue
- PRO/OMEGA: Priority support via email or Slack
๐ง Contact
- Support: support@narapallc.com
- General inquiries: contact@narapallc.com
- GitHub: @artvepa80
- Company: Narapa LLC, Miami, Florida
- Website: Coming soon
๐ License
MIT License for core functionality. See LICENSE for details.
PRO and OMEGA Guardian features are licensed separately under commercial terms.
๐ Acknowledgments
Built with:
- Google Gemini for AI analysis
- BigQuery for analytics
- Pub/Sub for real-time alerts
- Love from Miami โ๏ธ
โญ Star us on GitHub if Hefesto helped you catch bugs!
Hefesto: AI-powered code quality that caught 3 critical bugs in its own release. Now protecting your code.
ยฉ 2025 Narapa LLC. All rights reserved.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file hefesto_ai-4.3.0.tar.gz.
File metadata
- Download URL: hefesto_ai-4.3.0.tar.gz
- Upload date:
- Size: 150.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
2bd361b9443aa6ca85a41c4f497940d05189319b9e6584df25908e927e740692
|
|
| MD5 |
2ee74ce5afe8c55af0e88c7ff404577b
|
|
| BLAKE2b-256 |
8feab798485b29db1e36e298f41ce953f285381feb618cf399b3d0cceecc165c
|
File details
Details for the file hefesto_ai-4.3.0-py3-none-any.whl.
File metadata
- Download URL: hefesto_ai-4.3.0-py3-none-any.whl
- Upload date:
- Size: 158.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
a8ea5f42851914ff767ca6e578e11524ca7072646addb822d5726cb2d729db86
|
|
| MD5 |
983c4caa5f3d17789ae03b899eb191e1
|
|
| BLAKE2b-256 |
3378251a400f94f23c96ed299e8027af0d60021c605ae1ef09ea1d78d1facc13
|