HCP
Context Engineering Protocol
Enterprise
Enterprise Solutions
Pricing
Plans & Pricing
Open Source
Open Source Projects
HALETHEIA Security Stack v1.0
AI-Safe Development Standard
Cross-IDE Security Tools for AI Coding
The HALETHEIA Security Stack solves the 6 critical problems of AI coding. Validate dependencies, optimize costs, sandbox execution, and audit LLM decisions. Production-ready components for secure AI development.
✅ Sentinel v0.2.0 Production✅ aronly v0.24.0 Production🚧 babuino MVPMIT License
The 6 Critical Problems of AI Coding
Security challenges unique to AI-assisted development
1
Implicit Execution
AI runs code without explicit approval. No sandbox, no rollback.
babuinoSentinel
2
Unverifiable Trust
Packages suggested by LLMs are unverified. No registry check.
Sentinelcutufato
3
System Opacity
Zero visibility into LLM reasoning. Can't audit decisions.
cutufato
4
Exposure Time
New packages with 0 days of community scrutiny trusted instantly.
Sentinel
5
Attack Surface
AI agents with file + network access = infinite blast radius.
babuinoSentinel
6
Lack of Governance
No policies, no budgets, no accountability for AI-driven costs.
Sentinelaronly
The 4 Components
Each component addresses specific security challenges
Get Started (5 Minutes)
Setup the full security stack right now
1. Install Sentinel
pip install sentinel-deps[all]
cd my-project
sentinel init --lang python --policy balanced2. Install aronly
git clone https://github.com/haletheia/aronly
cd aronly && cargo install --path .
aronly init --global # Installs auto-rewrite hook3. Start Coding
Your workflow doesn't change. Security is transparent:
- Sentinel validates in real-time (via MCP)
- aronly optimizes automatically (via hook)
- babuino executes safely (CLI when needed)
4. Verify Savings
sentinel scan requirements.txt
aronly gain # See your token savingsUse Cases
Security Stack adapts to different needs
🚀 Startups (Cost-Conscious)
- aronly: Reduce OpenAI/Anthropic bills 70%
- Sentinel: Prevent vulnerabilities before production
- babuino: Execute third-party apps without risk
useCases.startups.footer
🏢 Enterprise (Security-First)
- Sentinel strict policy: CVE blocking + allowlist
- cutufato: Audit trails for compliance
- babuino: Mandatory sandbox for external code
useCases.enterprise.footer
👨💻 Freelancer (Productivity)
- aronly: More LLM queries with same budget
- Sentinel: Instant package validation (MCP)
- Zero friction: Workflow unchanged
useCases.freelancer.footer
Proven Results
Production metrics from real deployments
metrics.tokenSavings
metrics.tokenSavingsLabel
metrics.languagesSupported
metrics.languagesSupportedLabel
metrics.wasmStartup
metrics.wasmStartupLabel
metrics.idesSupported
metrics.idesSupportedLabel
Real-World Examples
Technical community members have documented their experiences
examples.example1.quote
A developer shared their 30-day analysis using aronly in a React/Node project, with detailed metrics showing $155/month savings per developer.
aronlyReal MetricsProduction
examples.example2.quote
An architect documented how Sentinel prevented critical vulnerabilities in their enterprise deployment, including 2 supply chain attacks.
SentinelEnterpriseCase Study
examples.example3.quote
A security researcher tested babuino with intentionally malicious prompts. The sandbox successfully contained all attempted exploits.
babuinoSecurity ResearchPenetration Test
💡 Tip: Search "Sentinel dependency firewall" or "aronly token optimization" to find more community content and detailed guides.
Works Seamlessly with HCP
Security Stack integrates naturally with Human-Code-AI Protocol
Complete Workflow
= The only complete solution for secure AI coding
Ready to secure your AI coding workflow?
Production-ready tools. Zero friction. Setup in 5 minutes.