ccglass

ccglass

ccglass is an MIT-licensed local reverse proxy and web dashboard that reveals exactly what coding agents send to their models. It sits between agents like Claude Code, Codex, DeepSeek-TUI, Kimi, OpenCode, Ollama, or OpenRouter and the API, streaming requests to a live dashboard with turn diffs, latency, cache, and cost.

free
reverse proxycoding agentdebuggingobservabilityClaude CodeCodexDeepSeekOllamaMCPopen source
Indexed
Updated
4.1 (0 Number of reviews)

Log in to rate the project

Try Now

What ccglass Is

ccglass is a lightweight local reverse proxy plus web dashboard for coding agents. Instead of guessing what an agent is doing inside its loop, you point the agent at ccglass and get a full view of every request and response, on your own machine. It supports Claude Code, Codex, DeepSeek-TUI, Reasonix, Kimi, OpenCode, Ollama, OpenRouter, and more.

What the Dashboard Shows

  • Live request stream: API calls captured as they happen, with per-request latency, token, cache, and cost data.
  • Conversation flow diagram: the agent loop laid out visually, plus turn-to-turn diffs to see what actually changed.

Working with the Data

Responses can be exported in HTTP, Markdown, JSON, or HAR format, and a session summary rolls up statistics across a run. MCP self-inspection lets you look at what an MCP server is receiving and returning. Light and dark themes are both available.

License

ccglass is MIT licensed, free and open source. It runs locally, so requests do not have to leave the developer machine to be inspected.

Pros & Cons

Pros

  • Runs locally, so agent traffic can be inspected without shipping it to a third party
  • Broad agent support out of the box, including Claude Code, Codex, DeepSeek-TUI, Kimi, OpenCode, Ollama, and OpenRouter
  • Turn-to-turn diffs, cost, cache, and latency in one dashboard
  • MIT licensed and free

Cons

  • Aimed at developers, not end users, and requires configuring an agent to route through the proxy
  • No hosted or team version is described on the site
  • Feature depth depends on the specific agent under test

Frequently Asked Questions

Which agents work with ccglass?

Claude Code, Codex, DeepSeek-TUI, Reasonix, Kimi, OpenCode, Ollama, OpenRouter, and more per the site.

Where does the data go?

ccglass runs locally as a reverse proxy plus dashboard, so nothing is shipped off the developer machine by default.

Can I export a response?

Yes, supported formats include HTTP, Markdown, JSON, and HAR.

How much does it cost?

It is MIT licensed and free.

Explore More

Similar Tools

StackBuilder

StackBuilder

StackBuilder is a free, AI-driven tool that generates professional cloud architecture diagrams from natural language descriptions. It supports major platforms like AWS, Azure, GCP, and Kubernetes, and allows exports to PNG, SVG, and PDF. No registration is required, making it ideal for system design interviews, architecture documentation, and presentations.

Nest by RAVEN

Nest by RAVEN

Nest by RAVEN (also known as NestMux) is a multi-AI terminal workbench for developers. It allows parallel execution of Claude, Gemini, Codex, Copilot, and Aider within a single window. Features include Git worktrees integration, team terminal sharing, MCP panel, and broadcast prompts. It's local-first, telemetry-free, and available for free download on macOS, Windows, and Linux.

h5i

h5i

h5i is an open-source sandbox tool designed for AI coding agents like Claude Code and Codex. It encapsulates the agent, shell, dependencies, and browser within a single, isolated boundary. This lightweight sandbox launches in under 200ms, supports exporting auditable patches and execution logs, and is local-first, SaaS-free, delivered as a single Rust binary.

Deep Work Plan

Deep Work Plan

Deep Work Plan is an open-source methodology that transforms any code repository into a structured, AI-executable environment using an `init.md` file. It breaks down long-term coding tasks into atomic steps with clear acceptance criteria, validation gates, and recoverable states, preventing AI agents from derailing. It's agent-agnostic, open-source (MIT), and prevents vendor lock-in.

Spanly

Spanly

Spanly offers a specialized observability and monitoring solution for Model Context Protocol (MCP) servers. It helps SaaS teams track error rates, session traces, latency, and client behavior in production environments. With quick CLI/SDK integration, it complements existing monitoring stacks and provides data residency options in the US and EU. A free scanner is available to quickly identify protocol-level vulnerabilities.

Yaeris

Yaeris

Yaeris is a marketplace for Model Context Protocol (MCP) servers, offering a centralized directory where developers can freely browse and publish human-reviewed MCP servers. It supports OAuth/API token integration, making it easy for AI agents to connect with real-world tools. While basic features are free, paid add-ons are available to boost server visibility. It's a pragmatic solution for AI agent developers and software companies looking to streamline their integrations.

Open-source Alternatives

guidellm: Open-Source Tool for Evaluating and Optimizing LLM Inference

guidellm is an open-source tool developed by the vLLM team to evaluate and optimize Large Language Model (LLM) inference performance in production environments. It offers stress testing, latency analysis, and throughput assessment to help developers identify bottlenecks and fine-tune deployment configurations. The project is primarily written in Python and licensed under Apache-2.0. At the time of collection, it had 1214 stars on GitHub.

ai-gateway: Unified AI Gateway Based on Envoy Gateway

ai-gateway is an open-source project built on Envoy Gateway, offering a unified API gateway to manage access to diverse generative AI services. It simplifies AI application integration and operations by providing features like load balancing, caching, and rate limiting for various AI providers. The project is written in Go and licensed under Apache-2.0.

Kun: Local-First AI Agent Workspace

Kun is a local-first AI agent workspace that unifies coding, writing, design, research, and automation through a shared GUI and TUI runtime. The project is primarily developed in TypeScript and has an 'Other' license. As of collection time, it has 4813 GitHub stars.

go-micro: Go framework fusing AI agent harness with microservices

go-micro is an open-source Go framework that fuses an AI agent harness with microservices, supporting MCP, A2A, and multi-LLM integration. It is licensed under Apache-2.0 and primarily written in Go. As of the collection time, the project had 22,755 stars on GitHub.

terax-ai: Lightweight Tauri-based Desktop Dev Environment

terax-ai is a Tauri-based desktop development environment with a size of only 7-8 MB. It integrates a GPU terminal, CodeMirror editor, Git tools, and multi-provider AI agents, offering an all-in-one development experience. The project is primarily written in TypeScript and licensed under Apache-2.0.

jar-analyzer: Open-Source GUI Tool for Java JAR Analysis with AI Assistant

jar-analyzer is an open-source GUI tool for Java JAR package analysis, featuring an integrated AI assistant. It offers robust capabilities like JAR DIFF, method call graph exploration, DFS call chain analysis, taint analysis, and control flow graph (CFG) program analysis. Ideal for Java developers and security researchers, it streamlines code auditing and reverse engineering tasks. The primary language is Java, licensed under GPL-3.0, with 2111 GitHub stars at the time of collection.