Standards Reference
Comprehensive index of universal standards accessible via search_standards.
Overview​
prAxIs OS ships two types of standards. The first are timeless CS fundamentals that apply across programming languages. The second are the standards that form the basis for prAxIs OS itself. These help aid the agents in making better decisions and adhering to standards served via the MCP search_standards tool. Standards are retrieved via semantic search (search_standards) rather than direct file reading.
Access Pattern:
search_standards("how to handle race conditions")
# Returns relevant chunks from universal standards
Standards Categories​
AI Assistant​
Standards for AI agent behavior and decision-making.
| Standard | Purpose |
|---|---|
| Agent Decision Protocol | Query-first approach, multi-angle thinking, probabilistic behavior management |
| prAxIs OS Orientation | 8 mandatory bootstrap queries for AI agents |
| prAxIs OS Development Process | Systematic development workflow patterns |
| Analysis Methodology | Problem analysis and decomposition strategies |
| Commit Protocol | Commit message standards and git workflows |
| Compliance Protocol | Ensuring adherence to project standards |
| Knowledge Compounding | Building cumulative understanding across sessions |
| MCP Tool Discovery | Finding and using MCP tools effectively |
| MCP Tools Guide | Comprehensive MCP tool usage patterns |
| Pre-Generation Validation | Validation before code generation |
| Query Construction Patterns | Effective search_standards query patterns |
| RAG Content Authoring | Writing discoverable standards content |
| Standards Creation Process | Creating new universal standards |
AI Safety​
Safety rules and protections for AI agents.
| Standard | Purpose |
|---|---|
| Credential File Protection | Preventing accidental credential exposure |
| Date Usage Policy | Using current_date tool vs cached dates |
| Git Safety Rules | Safe git operations (no force push, etc.) |
| Import Verification Rules | Verifying imports exist before using |
| Production Code Checklist | Pre-commit quality checklist |
Architecture​
Software architecture patterns and principles.
| Standard | Purpose |
|---|---|
| API Design Principles | RESTful design, versioning, error handling |
| Dependency Injection | Loose coupling, testability, flexibility |
| Separation of Concerns | Layered architecture, bounded contexts |
| SOLID Principles | Single Responsibility, Open/Closed, Liskov, Interface Segregation, Dependency Inversion |
Concurrency​
Patterns for managing shared state and parallel execution.
| Standard | Purpose |
|---|---|
| Deadlocks | Detection, prevention, resolution of deadlocks |
| Locking Strategies | Mutex, RWLock, fine-grained locking patterns |
| Race Conditions | Detection and prevention of race conditions |
| Shared State Analysis | Identifying and managing shared data |
Database​
Database design and implementation patterns.
| Standard | Purpose |
|---|---|
| Database Patterns | Transactions, indexes, migrations, query optimization |
Documentation​
Documentation standards and templates.
| Standard | Purpose |
|---|---|
| API Documentation | API docs structure and content |
| Code Comments | When and how to comment code |
| README Templates | Project README structure |
Failure Modes​
Resilience and graceful degradation patterns.
| Standard | Purpose |
|---|---|
| Circuit Breakers | Preventing cascading failures |
| Graceful Degradation | Degrade functionality, not availability |
| Retry Strategies | Exponential backoff, jitter, retry limits |
| Timeout Patterns | Preventing indefinite waiting |
Installation​
Installation and update procedures.
| Standard | Purpose |
|---|---|
| Gitignore Requirements | Required .gitignore entries for prAxIs OS |
| Update Procedures | Upgrading prAxIs OS safely |
Meta-Framework​
Principles for building AI-assisted workflows.
| Standard | Purpose |
|---|---|
| Command Language | Structured commands for AI execution |
| Framework Creation Principles | Creating new workflows |
| Horizontal Decomposition | Breaking tasks into optimal file sizes |
| Three-Tier Architecture | Execution, methodology, output file separation |
| Validation Gates | Evidence-based phase progression |
Performance​
Performance optimization patterns.
| Standard | Purpose |
|---|---|
| Optimization Patterns | Profiling, caching, lazy loading, algorithmic optimization |
Security​
Security patterns and best practices.
| Standard | Purpose |
|---|---|
| Security Patterns | Input validation, least privilege, defense in depth |
Testing​
Test strategies and patterns.
| Standard | Purpose |
|---|---|
| Integration Testing | Testing component interactions |
| Property-Based Testing | Automated test case generation |
| Test Doubles | Mocks, stubs, fakes, spies |
| Test Pyramid | Unit (70%), Integration (20%), E2E (10%) |
Workflows​
Workflow system standards.
| Standard | Purpose |
|---|---|
| MCP RAG Configuration | Configuring RAG for workflow discovery |
| Time Estimation Standards | Estimating workflow and task durations |
| Workflow Construction Standards | Building compliant workflows |
| Workflow Metadata Standards | metadata.json structure and requirements |
| Workflow System Overview | How the workflow system works |
Usage Pattern​
Standards are accessed via semantic search, not direct file reading:
# Query for specific guidance
search_standards(
query="How do I prevent race conditions in async code?",
n_results=3
)
# Returns relevant chunks:
# - Universal race condition principles
# - Language-specific patterns (if generated)
# - Project-specific guidance
Benefits:
- 90% context reduction per query (2-5KB chunks vs 50KB files)
- Only relevant content loaded
- Maintains attention quality
- Just-in-time delivery drives query-first behavior
- Results in 71% fewer messages overall
Related Documentation​
- MCP Tools Reference - Complete
search_standardsAPI documentation - How It Works - RAG-driven behavioral reinforcement
- Architecture - MCP/RAG system implementation
- Workflows - How standards integrate with workflows