RobThePCGuy/Claude-Patent-Creator — USPTO patent creation system with MCP server + Claude Code plugin.

Description and use

**Brief summary (2‑3 sentences)**
RobThePCGuy/Claude‑Patent‑Creator is an open‑source USPTO‑style patent drafting platform that couples a Model‑Context‑Protocol (MCP) server with a Claude code‑plugin. It offers hybrid RAG over the MPEP, USC, CFR and a BigQuery table of > 76 M patents, runs automated 35 USC 112 compliance checks, performs prior‑art searches, and can generate claim‑diagrams on GPU‑accelerated agents.

**Value**
The project turns a generic LLM assistant into a domain‑specific patent‑authoring engine by exposing real‑world tools (search, compliance, diagramming) through a standard MCP API. This lets any AI agent—Claude, ChatGPT, or custom models—invoke authoritative legal data, run deterministic checks, and retrieve results in a uniform way, dramatically reducing the engineering effort required to build a compliant, data‑rich patent‑drafting workflow.

**Practical adoption path**

| Step | Action | Reason |
|------|--------|--------|
| 1️⃣ Evaluate the MCP/CLI | Clone the repo, run the provided Docker‑compose or pip install and fire up the MCP server; use the sample Python SDK to issue a simple “search prior art” request. | Confirms that the API contract matches your existing AI orchestration layer. |
| 2️⃣ Connect your LLM agent | Point your Claude/ChatGPT wrapper to the MCP endpoint; register the Claude‑Code plugin if you need code‑generation assistance. | Enables seamless tool calls from the same prompt flow you already use. |
| 3️⃣ Pilot a specific use case | Start with a low‑risk task such as automated 112‑compliance checking on a set of draft claims. Capture logs, latency, and accuracy. | Provides a measurable ROI and surface‑level validation before scaling. |
| 4️⃣ Extend the pipeline | Add custom agents (e.g., a citation‑formatter or a diagram‑renderer) via the MCP SDK; optionally hook the BigQuery patent table for bulk analytics. | Leverages the platform’s extensibility to cover the full patent‑drafting lifecycle. |
| 5️⃣ Deploy to production | Containerize the MCP server, enable GPU instances for diagram generation, and enforce API‑key authentication. | Aligns with the project’s production‑ready signals (active maintainer, recent commits, 143 stars). |

**Production readiness**

• **Activity & community** – 143 GitHub stars, 24 forks, last commit 2026‑06‑25, and a Python‑centric codebase make it easy to audit and contribute.
• **Architecture** – The MCP server, CLI, and SDK are clearly defined; GPU acceleration and BigQuery integration are already packaged, reducing infrastructure glue code.
• **Reliability** – Automated 35 USC 112 checks and RAG over authoritative sources provide deterministic outputs, a key requirement for legal workflows.
• **Risk considerations** – Licensing and security posture still need a formal review, but no glaring metadata issues were found. Overall, the project is “high” for an OSS pilot and can be moved into production after the standard security and compliance vetting.

Problem

RobThePCGuy/Claude-Patent-Creator helps connect AI assistants to real tools and data through a standard protocol.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• connect AI agents to tools
• ship Model Context Protocol servers
• standardize integrations

Evidence

USPTO patent creation system with MCP server + Claude Code plugin. Hybrid RAG search over MPEP/USC/CFR, BigQuery access to 76M+ patents, automated 35 USC 112 compliance checks, prior art search, diagram generation. GPU-accelerated with skills and autonomous agents.

Source