Skip to content

a5c-ai/babysitter / library/specializations/ai-agents-conversational/skills/mcp-csp-investigation

mcp-csp-investigation

Verified

Comprehensive Content Security Policy audit for MCP Apps in sandboxed iframes. Discovers all network origins, traces them to source, and generates CSP configuration for registerAppResource.

FileKindSize
README.mdmarkdown2.4 KB
SKILL.mdskill md9.2 KB
# MCP CSP Investigation Skill

## Overview

The `mcp-csp-investigation` skill performs comprehensive Content Security Policy audits for MCP Apps running in sandboxed iframes. This is critical because ALL network requests fail **silently** in sandboxed iframes without proper CSP declarations -- no errors, no warnings, just silent failure.

## Key Features

- **Build Output Analysis**: Discovers every network origin in compiled HTML/CSS/JS
- **Origin Tracing**: Traces each origin to its source (constant, env var, conditional)
- **Third-Party Detection**: Finds hidden network requests in dependencies
- **CSP Generation**: Produces categorized domain lists (resource, connect, frame)
- **Environment Annotation**: Tags origins as universal, dev-only, or prod-only

## Prerequisites

1. **Build toolchain**: Ability to build the target application
2. **grep/ripgrep**: For searching build output
3. **Understanding of CSP**: Basic knowledge of Content Security Policy

## Quick Start

### 1. Build the Application

```bash
npm run build
```

### 2. Search for Network Origins

```bash
grep -rEoh 'https?://[a-zA-Z0-9._-]+' dist/ | sort -u
```

### 3. Categorize and Configure

Place CSP in `contents[]` of `registerAppResource` read callback:

```typescript
contents: [{
  uri: 'app:///my-app',
  mimeType: RESOURCE_MIME_TYPE,
  text: bundledHtml,
  resourceDomains: ['https://cdn.example.com'],
  connectDomains: ['https://api.example.com'],
  frameDomains: [],
}]
```

## Why This Matters

MCP Apps run in sandboxed iframes with `sandbox="allow-scripts"`. There is no same-origin server. If you load a font from Google Fonts, fetch data from an API, or use a CDN for a library -- **all of these will silently fail** unless declared in CSP. The app will appear broken with no indication of why.

## Integration with Babysitter Processes

| Process | Integration Point |
|---------|------------------|
| convert-web-app-to-mcp.js | Phase 3: CSP investigation (critical security step) |
| migrate-openai-app-to-mcp.js | Phase 2: CSP investigation for existing OpenAI app |

## References

- [MCP Apps CSP/CORS Guide](https://github.com/modelcontextprotocol/ext-apps/blob/main/docs/csp-cors.md)
- [MDN Content Security Policy](https://developer.mozilla.org/en-US/docs/Web/HTTP/CSP)

## Related Resources

- mcp-app-verification - Validates CSP correctness post-configuration
- csp-security-auditor agent - Automated CSP auditing
rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics.Stars: 472kUpdated Claude CodeCodexVerifiedInstall
ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first.Stars: 194kUpdated Claude CodeCodexVerifiedInstall