Is the CapiscIO CLI really free forever?

Yes! The CLI is completely free and open source. No credit card required, no usage limits, no hidden fees. Use it in development, CI/CD, production monitoring—anywhere you need agent validation.

How do I install the CapiscIO CLI?

Install via npm with "npm install -g capiscio-cli" or pip with "pip install capiscio-cli". The CLI works on macOS, Linux, and Windows. Full installation docs available at docs.capiscio.com/cli/installation.

Can I use the CLI to monitor third-party agents?

Absolutely! The CLI is perfect for continuous monitoring of partner agents. Set up cron jobs or scheduled tasks to validate third-party agents on a schedule and get alerts when they fail validation.

Does the CLI work in CI/CD pipelines?

Yes! The CLI outputs structured JSON and uses standard exit codes (0 for pass, 1 for fail), making it ideal for GitHub Actions, GitLab CI, Jenkins, or any CI/CD system. We provide ready-to-use workflow templates.

What does the CLI validate?

The CLI performs comprehensive validation: Agent Card JSON schema, protocol compliance, endpoint availability, JWS signature verification, transport protocol support (REST/gRPC/MCP), and capability definitions. It provides three-dimensional scoring: Compliance, Trust, and Availability.

Can I customize validation rules?

Yes! Use configuration files to set custom thresholds, enable/disable specific checks, choose which transport protocols to test, and define failure criteria. Perfect for different environments (dev vs prod) or compliance requirements.

v2.0.0 • Three-Dimensional Scoring

Validate Agents Everywhere

Stop Debugging A2A Failures at 2 AM

Pre-deployment validation, CI/CD gates, third-party monitoring, and continuous health checks—all from one CLI.

�️

Pre-Deployment

Catch errors before commit

🔄

CI/CD Gates

Block invalid deployments

🔍

3rd Party Validation

Verify external agents

⏰

Health Monitoring

Continuous uptime checks

Get Started in 30 Seconds

Choose your path

Node.js (Recommended)

npm install -g capiscio-cli

Python

pip install capiscio

Other installation options →

Standalone Binary: GitHub Releases

GitHub Action: capiscio/validate-a2a@v1

Then validate:

capiscio validate ./agent.json

View on GitHub

Stop These Developer Nightmares

We've been there. We built this so you don't have to.

🔥

"It worked on my machine..."

— Every developer at 2 AM

You committed an Agent Card with a typo in the transport URL. CI passed. Production broke. Three hours debugging why requests fail.

✓ Solution:

capiscio validate ./agent.json

Catches errors pre-commit

👻

"Their agent stopped responding..."

— Platform engineer monitoring 3rd parties

You depend on 12 external agents. One changed their schema. Your users are complaining. You have no monitoring for A2A compliance.

✓ Solution:

capiscio validate https://3rdparty.com

Continuous monitoring + alerts

⏰

"CI says pass, prod says fail"

— DevOps lead with broken pipeline

Your tests check JSON validity. They don't check if the endpoint actually works, if signatures verify, or if transport configs are valid.

✓ Solution:

uses: capiscio/validate-a2a@v1

Live endpoint + signature checks

Four Ways to Validate

Choose the workflow that fits your needs

🛠️

1. Pre-Deployment Validation

Catch errors before they reach Git

Install once:

npm install -g capiscio-cli

Validate anytime:

capiscio validate ./agent.json

Perfect for:

• Local development workflow
• Pre-commit validation hooks
• Quick agent card debugging

🔄

2. CI/CD Quality Gate

Block invalid agents from deploying

Add to workflow:

- uses: capiscio/validate-a2a@v1 with: agent-card: ./agent.json strict: true

Perfect for:

• GitHub Actions / GitLab CI
• Pull request validation
• Deployment gating with scores

🔍

3. Third-Party Agent Validation

Verify external agents you depend on

Validate any live agent:

capiscio validate https://external-agent.com

Get trust scores:

Compliance: 100/100
Trust: 90/100 (Valid signature)
Availability: 95/100

Perfect for:

• Vetting partner agents
• Registry submissions
• Due diligence on new integrations

⏰

4. Continuous Health Monitoring

Schedule recurring validation checks

Cron job example:

# Every 6 hours0 */6 * * * capiscio validate \https://agent.com --json \> /var/log/agent-health.log

GitHub Actions schedule:

on: schedule: - cron: '0 */6 * * *'

Perfect for:

• Uptime monitoring of your agents
• Watching 3rd party agent health
• Alerting on compliance degradation

Why Choose CapiscIO CLI?

Production-ready validation for AI agents

�

NEW v2.0

Three-Dimensional Scoring

• Compliance score (0-100)
• Trust score with signature confidence
• Availability score from live testing
• Production readiness assessment

🔒

Cryptographic Trust

• JWS signature validation (RFC 7515)
• JWKS endpoint verification
• Trust confidence multiplier
• Secure by default architecture

Live Transport Testing

• JSONRPC, GRPC, REST validation
• Real connectivity checks
• Cross-transport consistency
• Availability scoring

A2A Protocol Expert

• Official A2A v0.3.0 compliance
• Legacy format support
• GitHub Action available

Usage Examples

Common validation scenarios

Basic Agent Validation

Validate agents from URLs or local files

# Works with both npm and pip installations

capiscio validate https://your-agent.com

# Validate local file

capiscio validate ./agent-card.json

# Auto-detect in current directory

capiscio validate

💡 Installation Options:

• npm: npm install -g capiscio-cli

• pip: pip install capiscio

• Binary: Download from GitHub releases

Validation Output

Clear, actionable feedback with three-dimensional scoring

Successful Validation

✅ A2A AGENT VALIDATION PASSED

Agent: https://api.example.com

Version: 0.3.0 (Strictness: progressive)

📊 THREE-DIMENSIONAL SCORES:
• Compliance: 100/100 (A+)
• Trust: 90/100 (A) - Valid signature, 1.0x confidence
• Availability: 95/100 (A+) - All endpoints responsive

🔍 VALIDATION SUMMARY:
📊 12 checks performed: 12 passed, 0 failed, 0 warnings
⏱️ Completed in 245ms

🏆 Production Ready! All criteria met.
🚀 Your agent is ready for deployment!

Failed Validation

❌ A2A AGENT VALIDATION FAILED

Agent: ./agent-card.json

Version: 0.3.0 (Strictness: progressive)

📊 THREE-DIMENSIONAL SCORES:
• Compliance: 85/100 (B)
• Trust: 40/100 (D) - No signature, 0.6x confidence
• Availability: N/A - Endpoint unreachable

🔍 VALIDATION SUMMARY:
📊 12 checks performed: 8 passed, 4 failed, 2 warnings
⏱️ Completed in 89ms

🔍 ERRORS FOUND (2):
❌ ENDPOINT_UNREACHABLE: Primary endpoint not accessible
❌ MISSING_SECURITY_SCHEME: No security scheme defined

⚠️ NOT PRODUCTION READY
Compliance ≥95, Trust ≥60, Availability ≥80 required

Perfect For

Built for the AI agent ecosystem

AI/LLM Developers

Building agent-based applications

Validate your AI agents before deployment and ensure they meet A2A protocol standards.

DevOps Teams

Validating in CI/CD pipelines

Integrate validation into your deployment pipeline with JSON output and error codes.

Registry Operators

Ensuring protocol compliance

Validate agent submissions and ensure registry quality with comprehensive checks.

AI Researchers

Agent-to-agent communication

Ensure your research agents conform to interoperability standards.

Product Teams

Deploying AI agents to production

Ship AI agents with confidence using production-ready validation.

Platform Builders

Creating agent ecosystems

Build platforms that require validated, compliant AI agents.

Frequently Asked Questions

Everything you need to know

Runtime Validation

Need Real-Time Production Validation?

The CLI is perfect for development and CI/CD, but for live production monitoring, you need CapiscIO Middleware.

Runtime validation - Catch issues as they happen in production
Zero code changes - Drop-in middleware for Python
Performance metrics - Track validation speed and compliance

Explore Middleware

# Python middleware in action

from capiscio_a2a_security import secure, SecurityConfig

agent = secure(
    MyAgentExecutor(),
    SecurityConfig.production()
)

# ✅ Every request validated
# ⚡ <2ms overhead
# 🔒 100% compliant

Ready to Validate Your AI Agents?

Join the growing ecosystem of A2A protocol compliant AI agents

npm: npm install -g capiscio-cli && capiscio validate

pip: pip install capiscio && capiscio validate

View Documentation

MIT Licensed • Zero Dependencies • Production Ready