Charles-mcp

charles-mcp is an MCP server that connects an AI agent to Charles Proxy’s Web Interface/recording data, enabling live capture reading and structured analysis of HTTP traffic (with a summary-first approach and drill-down to per-entry details). It can also analyze recorded sessions/history and provide control/status tools for the Charles environment.

Evaluated Mar 30, 2026 (0d ago)
Repo ↗ DevTools mcp charles-proxy traffic-analysis network-debugging python stdio-mcp live-capture http debugging-tools
⚙ Agent Friendliness
58
/ 100
Can an agent use this?
🔒 Security
29
/ 100
Is it safe for agents?
⚡ Reliability
41
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
78
Documentation
75
Error Messages
--
Auth Simplicity
30
Rate Limits
10

🔒 Security

TLS Enforcement
20
Auth Strength
35
Scope Granularity
10
Dep. Hygiene
60
Secret Handling
25

Tooling is aimed at local proxy diagnostics but the provided README states that tools return complete raw original content by default (risk of exposing secrets/tokens from traffic). Charles Web Interface credentials are passed via environment variables and the README includes default credentials (admin/123456), which increases the risk if not changed. No explicit mention of TLS requirements or fine-grained authorization/scope controls for the MCP server.

⚡ Reliability

Uptime/SLA
0
Version Stability
55
Breaking Changes
45
Error Recovery
65
AF Security Reliability

Best When

You are running Charles locally with the Web Interface enabled and want an MCP-enabled agent to summarize and explore traffic efficiently without receiving full raw packet bodies by default.

Avoid When

You cannot secure access to Charles Web Interface credentials or you cannot prevent agents from viewing sensitive headers/bodies contained in captured traffic.

Use Cases

  • Real-time debugging of web traffic while a developer/tester watches Charles sessions
  • AI-assisted triage of “what just happened” by grouping and summarizing live requests
  • Investigating specific requests/responses by drilling into an entry’s detail when a summary indicates a target
  • Analyzing historical Charles recordings (.chlsj) with structured summaries and match reasoning
  • Network condition simulation via Charles throttling controls during agent workflows

Not For

  • Production-grade secure data processing without access controls (it is designed for local proxy/diagnostics workflows)
  • Handling sensitive data without an explicit redaction/masking layer (tooling states it returns full raw content by default)
  • High-assurance compliance use cases where audit requirements and formal contracts are mandatory

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: Basic authentication to Charles Web Interface using CHARLES_USER/CHARLES_PASS (via MCP process environment)
OAuth: No Scopes: No

Authentication is delegated to the Charles Web Interface. MCP server auth itself is not described; instead it relies on Charles credentials passed via env vars.

Pricing

Free tier: No
Requires CC: No

No pricing information provided; repository/manifest indicates an MIT-licensed Python package intended for self-hosting.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Documented

Known Gotchas

  • Default tool behavior returns raw captured content (including sensitive headers/bodies) unless downstream masking is applied.
  • Large detail responses may exceed context; tool warns and supports limiting include_full_body/max_body_chars.
  • Live stop can return stop_failed; agent should check recoverable/active_capture_preserved to decide next steps.
  • query_live_capture_entries is designed not to advance cursor; repeated calls should continue to work rather than “consume” data.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for Charles-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

6523
Packages Evaluated
19880
Need Evaluation
586
Need Re-evaluation
Community Powered