home / skills / a5c-ai / babysitter / request-debugger

This skill provides comprehensive HTTP request/response debugging and inspection, automatically redacting sensitive data and exporting reproducible curl

npx playbooks add skill a5c-ai/babysitter --skill request-debugger

Review the files below or copy the command above to add this skill to your agents.

Files (2)
SKILL.md
1.9 KB
---
name: request-debugger
description: HTTP request/response debugging and inspection tools
allowed-tools:
  - Read
  - Write
  - Edit
  - Glob
  - Grep
  - Bash
---

# Request Debugger Skill

## Overview

This skill implements comprehensive HTTP request/response debugging and inspection tools for SDKs, helping developers understand and troubleshoot API interactions.

## Capabilities

- Log request/response details with formatting
- Redact sensitive data automatically
- Correlate logs with request IDs
- Support curl command export for reproduction
- Implement request timing breakdowns
- Capture retry attempts and failures
- Support HAR (HTTP Archive) export
- Enable request/response interception

## Target Processes

- Error Handling and Debugging Support
- Logging and Diagnostics
- Developer Experience Optimization

## Integration Points

- HTTP client interceptors
- Logging frameworks
- Debug proxies (Charles, mitmproxy)
- Browser DevTools protocols
- IDE debug integration

## Input Requirements

- Redaction rules for sensitive data
- Logging level configurations
- Timing breakdown requirements
- Export format preferences
- Interception capabilities

## Output Artifacts

- Debug interceptor middleware
- Request logger implementation
- Sensitive data redactor
- Curl command generator
- HAR exporter
- Timing breakdown utilities

## Usage Example

```yaml
skill:
  name: request-debugger
  context:
    logging:
      enabled: true
      level: debug
      prettyPrint: true
    redaction:
      headers: ["Authorization", "X-Api-Key"]
      body: ["password", "secret", "token"]
    features:
      curlExport: true
      harExport: true
      timingBreakdown: true
      retryLogging: true
    requestIdHeader: "X-Request-ID"
```

## Best Practices

1. Always redact sensitive data
2. Include timing breakdowns
3. Support reproducible curl export
4. Log retry attempts with reasons
5. Use structured logging format
6. Enable selective debug activation

Overview

This skill provides HTTP request and response debugging and inspection tools tailored for SDKs and developer workflows. It helps capture, format, and export detailed request data while protecting sensitive information. The toolset focuses on reproducibility, timing analysis, and clear correlation between requests and logs.

How this skill works

The skill installs as HTTP client interceptors or middleware to capture requests, responses, retries, and timing breakdowns. It automatically redacts configured sensitive fields, correlates entries by request ID, and can export reproductions as curl commands or HAR files. Developers configure logging level, redaction rules, and export preferences to fit their environment.

When to use it

  • Troubleshooting failed API calls with incomplete server errors
  • Analyzing request latency and identifying slow network or service components
  • Recording reproducible requests for bug reports or support
  • Auditing and debugging retry behavior and failure reasons
  • Integrating richer diagnostics into SDKs or developer tooling

Best practices

  • Always configure redaction rules for headers and body fields before enabling debug logs
  • Enable timing breakdowns to separate DNS, TCP, TLS, and server processing times
  • Use structured logs and include the request ID to make traces searchable
  • Export curl or HAR when sharing issues with external teams to reproduce reliably
  • Enable selective debug activation to avoid logging sensitive production traffic

Example use cases

  • Attach as an HTTP interceptor in an SDK to capture request/response cycles and retries
  • Generate curl commands from captured requests to reproduce issues locally
  • Export a session as a HAR file for postmortem analysis or integration with debugging proxies
  • Log timing breakdowns to identify whether latency is client-side, network, or server-side
  • Mask API keys and passwords automatically while keeping helpful debugging metadata

FAQ

How does redaction work?

You supply rules listing headers and body fields to redact. The skill replaces configured values with sanitized placeholders before logging or exporting.

Can I export captured traffic for external tools?

Yes — the skill can generate curl commands and export HAR files to reproduce requests in other debugging tools or for sharing with support teams.

Will this impact production performance?

Interceptors add minimal overhead but timing collection and detailed logging have cost. Use selective activation and sampling in production to reduce impact.