heypandax/cc-pocket

Description and use

**Brief Summary (2‑3 sentences)**
heypandax/cc‑pocket is an open‑source Kotlin mobile client that lets you drive Claude Code or OpenAI Codex from a phone, with features such as session resumption, real‑time output streaming, and remote approval of tool permissions. The communication channel is end‑to‑end encrypted and operates as a zero‑knowledge relay, making it suitable for secure, on‑the‑go AI interactions.

**Value**

• **Fast AI enablement** – Developers can add generative‑code capabilities to prototypes or internal tools without building a model stack from scratch.
• **Secure mobile access** – End‑to‑end encryption and zero‑knowledge relaying protect prompts, code, and model responses, which is critical for confidential or regulated environments.
• **Iterative workflow support** – Session persistence and remote permission approval streamline RAG/agent experiments and allow quick debugging from anywhere.

**Practical Adoption Path**

1. **Clone & build** – Pull the repository, run the Gradle build, and generate a signed APK (or use the provided CLI wrapper).
2. **Configure credentials** – Insert your Claude Code or OpenAI API keys into the encrypted config file; the app will handle token rotation automatically.
3. **Integrate** – Use the exposed SDK/CLI to start a session from your backend, then connect the mobile client for interactive control, or embed the client in a CI/CD pipeline for remote approvals.
4. **Validate** – Run the built‑in test suite, verify the encrypted relay (e.g., using Wireshark) and confirm that no plaintext data leaves the device.

**Production Readiness**

• **Maturity** – Medium. The project is actively maintained (last update 2026‑06‑25) and has a modest but growing community (32 stars, 1 fork).
• **Security** – Strong design (E2E encryption, zero‑knowledge relay) but the license and long‑term maintainer commitment need a final review.
• **Operational considerations** – Verify dependency updates (Kotlin/Gradle libraries), assess the relay server’s scalability, and implement monitoring for API usage limits.
• **Suitable use cases** – Prototyping AI‑augmented features, internal tooling, or low‑risk production workflows after a short hardening cycle; not yet recommended for high‑throughput, customer‑facing services without further load testing and support contracts.

Use in a workflow

heypandax/cc-pocket helps add AI capability without starting from a blank model stack.

• prototype AI features
• build RAG or agent workflows
• evaluate model tooling

Limitations and review notes

Medium: useful for prototypes or internal workflows, with dependency and maintenance checks before production.