Add AgentAudit safe badge

[starbuck100]

AgentAudit Security Audit

Result: ✅ Safe (risk score: 0/100)

An automated security audit was performed on this package by AgentAudit. The audit followed a 3-pass process (Understand → Detect → Classify) and found no security findings.

What was checked

• Command injection patterns (exec, eval, subprocess)
• Credential theft / data exfiltration
• Obfuscation / backdoors
• MCP-specific risks (tool poisoning, path traversal, prompt injection)
• Supply chain risks
• Social engineering patterns

Report

Full report: https://agentaudit.dev/skills/mcp-context-forge

This PR adds an AgentAudit badge to the README indicating the package passed the security audit.

[crivetimihai]

Thanks for adding this, @starbuck100. LGTM — the badge placement is correct (below the tagline, above the banner image) and consistent with the other badges. The base64-encoded SVG in the shield URL avoids an external image dependency, which is a nice touch.