oasf

🥈Silver

Open Agentic Schema Framework (OASF) defines and manages AI agent capabilities, interactions, and metadata. It benefits operations teams by standardizing agent attributes and capabilities. OASF connects to AI agent workflows, enabling better integration and management of AI agents.

298340Updated Today

Intermediate30min to implementautomation

Saves ~240 min per use

Quick InstallView Source

git clone https://github.com/agntcy/oasf.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Define the agent's core attributes and capabilities. Use the [AGENT_NAME], [PRIMARY_FUNCTION], and [CAPABILITIES] placeholders to specify what the agent will do, its primary function, and the operations it supports.","tip":"Be specific about the input and output data structures. For example, if the agent processes customer data, define the exact fields required (e.g., customer_id, transaction_date, amount)."},{"step":"Specify the input and output schemas in detail. Use the [INPUT_SCHEMA] and [OUTPUT_SCHEMA] placeholders to define the data structure the agent expects to receive and the format it will return. Include validation rules, required fields, and data types.","tip":"Use tools like JSON Schema validators or Postman to test your schemas before finalizing them. Ensure all required fields are marked and data types are correctly specified."},{"step":"Configure metadata and security settings. Populate the [METADATA] placeholder with agent metadata such as version, author, license, and security protocols (e.g., authentication requirements, allowed roles).","tip":"Include security best practices like data retention policies, authentication methods, and role-based access control (RBAC) to ensure the agent operates securely in your environment."},{"step":"Define error handling and interaction protocols. Use the error_handling and interaction_protocols sections to specify how the agent should handle errors, retry failed operations, and interact with other systems (e.g., REST, gRPC).","tip":"Document fallback mechanisms and rate limits to ensure the agent behaves predictably under load or in error conditions. Test these scenarios in a sandbox environment."},{"step":"Validate and test the schema. Use OASF-compliant tools or validators to ensure the schema is valid and adheres to the OASF standards. Deploy the schema to your agent management platform.","tip":"Leverage tools like OASF CLI, Postman, or custom scripts to validate the schema. Test the agent in a staging environment to confirm it meets operational requirements."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/agntcy/oasf

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Generate an Open Agentic Schema Framework (OASF) compliant schema for an AI agent with the following requirements: [AGENT_NAME], [PRIMARY_FUNCTION], [INPUT_SCHEMA], [OUTPUT_SCHEMA], [CAPABILITIES], and [METADATA]. Ensure the schema includes versioning, security protocols, and error handling specifications. Format the output as valid JSON.

Example Output

```json
{
  "schema_version": "1.2.0",
  "agent_metadata": {
    "name": "DataProcessor_Agent",
    "description": "An AI agent specialized in cleaning, transforming, and validating structured data for analytics pipelines.",
    "version": "2.1.4",
    "author": "Analytics Team",
    "license": "MIT",
    "tags": ["data-processing", "ETL", "validation"],
    "security": {
      "auth_required": true,
      "allowed_roles": ["data-engineer", "analyst"],
      "data_retention_policy": "30 days"
    }
  },
  "capabilities": {
    "primary_function": "data_transformation",
    "supported_operations": ["cleaning", "normalization", "validation", "enrichment"],
    "input_schema": {
      "type": "object",
      "properties": {
        "raw_data": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "id": {"type": "string", "format": "uuid"},
              "timestamp": {"type": "string", "format": "date-time"},
              "metrics": {"type": "array", "items": {"type": "number"}},
              "metadata": {"type": "object"}
            },
            "required": ["id", "timestamp"]
          }
        },
        "config": {
          "type": "object",
          "properties": {
            "clean_missing_values": {"type": "boolean"},
            "normalize_units": {"type": "boolean"},
            "validation_rules": {"type": "array", "items": {"type": "string"}}
          }
        }
      },
      "required": ["raw_data"]
    },
    "output_schema": {
      "type": "object",
      "properties": {
        "processed_data": {
          "type": "array",
          "items": {
            "type": "object",
            "properties": {
              "id": {"type": "string", "format": "uuid"},
              "cleaned_metrics": {"type": "array", "items": {"type": "number"}},
              "validation_errors": {"type": "array", "items": {"type": "string"}},
              "processing_time": {"type": "string", "format": "date-time"}
            },
            "required": ["id", "cleaned_metrics"]
          }
        },
        "summary": {
          "type": "object",
          "properties": {
            "total_records": {"type": "integer"},
            "records_with_errors": {"type": "integer"},
            "processing_duration_ms": {"type": "integer"}
          }
        }
      },
      "required": ["processed_data", "summary"]
    },
    "error_handling": {
      "retry_policy": {
        "max_attempts": 3,
        "backoff_strategy": "exponential"
      },
      "fallback_mechanisms": ["log_errors", "notify_admin"],
      "error_schema": {
        "type": "object",
        "properties": {
          "error_type": {"type": "string"},
          "error_message": {"type": "string"},
          "timestamp": {"type": "string", "format": "date-time"},
          "affected_records": {"type": "array", "items": {"type": "string"}}
        }
      }
    },
    "interaction_protocols": {
      "supported_formats": ["REST", "gRPC", "WebSocket"],
      "rate_limits": {
        "max_requests_per_minute": 100,
        "burst_capacity": 200
      }
    }
  }
}
```