Open Context Standards

!Open Context Standards Logo

Initiative Charter - Get Involved

Mission

Open Context Standards is a community-led initiative that establishes and maintains open standards for how context for AI systems is structured, parsed, delivered, and verified across all computing platforms and AI systems. The aim to create a unified foundation that enables seamless interoperability between agents, models, and applications while reducing implementation overhead for developers and end users.

Why This Matters

The challenge: The rapid proliferation of AI agents, language models, and intelligent applications has created a fragmented landscape where each system handles context differently. For example, there are over 15 different ways for context to be delivered to developer tools today. This fragmentation leads to:

• Integration overhead - Developers must learn unique context handling and limitations for each tool
• Unpredictable behavior - No standardized expectations for how systems will process context
• Limited scalability - Difficulty scaling solutions across diverse AI ecosystems
• Format ambiguity - Uncertainty about supported context formats and structures

Our Solution: By establishing clear, open standards and practices, we create:

• Predictable implementations - Developers know what to expect from compliant systems
• Reduced development time - Common patterns eliminate the need to reinvent context handling and delivery
• Better tooling - Standardized formats enable better development and debugging tools
• Community alignment - Centralized discussion and decision-making around context practices

What we do

We identify emerging patterns, collaborate on opportunities where the community is getting value from emerging patters, help standardize practices, and drive adoption.

Our concern for context handling and delivery spans:

• Web platforms - Browser-based applications and services
• Mobile - iOS, Android, and cross-platform mobile apps
• Desktop applications - Native and cross-platform desktop software
• AI/ML systems - Large language models, agents, and AI-powered tools
• Developer toolchains - IDEs, build systems, and development workflows

These practices and standards address critical aspects of context handling and deliver such as:

• Parsing and Structure - How context data is organized and interpreted
• Delivery Mechanisms - Methods for finding and transmitting context between systems
• Verification Protocols - Ensuring context integrity and authenticity
• Format Specifications - Standardized representations for context data

How this relates to existing standards

Communication protocols such as MCP, A2A, HTTP, etc. all have their own standards and process for bringing the community together to solve problems. However, these systems are all finding, sharing, and passing along chunks of context and data between various systems and there’s no existing initiative to help with context itself. As a community, if we want to share ideas around how we could optimizing context delivery or formats, privacy preservations, etc. where would we go today? When a system produces context for an agent to incorporate into its system, what expectations should the system have with what the agent will do or understand with that context. This is where the Open Context Standards initiative is looking to play an important role of helping us advance context practices and standards.

Getting Involved

We welcome contributions from:

• Developers building context-aware applications
• Researchers working on context understanding and processing
• Tool Maintainers creating development and deployment tools
• Standards Organizations interested in context interoperability
• Industry Representatives from companies building AI and intelligent systems

Together, we'll create the foundation for the next generation of context-aware computing systems. Check out CONTRIBUTING.md for more details.