PLwC Personal

What PLwC Personal does

PLwC Personal connects compatible AI assistants to selected local capabilities through a gateway. It can read and organise files within bounded workspaces, create and read documents, maintain profiles, and retain preferences or working context across sessions and compatible clients. The system also includes reflection and Compile functions, plus a Governor workflow that keeps proposed persistent profile changes separate from their confirmed application.

The gateway exposes eight tools: plwc_status, plwc_describe, plwc_profile, plwc_reflection, plwc_governor, plwc_sandbox_run, plwc_workspace_operation, and plwc_document_operation. Document operations cover DOCX, XLSX, PPTX, PDF, and ZIP files. Python and Node.js execution is performed in separate Docker runtimes rather than through an automatic host-shell fallback.

How it works

PLwC Personal uses a local gateway as the MCP endpoint. Native STDIO integration is available for compatible MCP clients. A separate local Chat Bridge can connect ChatGPT in supported Chrome and Edge browser setups; the bridge has its own release version and protocol version.

Workspace actions are limited by locally configured boundaries. Document processing, Python execution, and Node.js execution use separate Linux-based Docker environments with network access disabled. The AI client and model remain your choice, but requests and returned content may still be processed by an external AI provider.

Persistent profile updates follow a Plan/Apply model. The Governor can separate a suggested lasting change from the later, explicit confirmation needed to apply it. This is useful when an assistant should be able to propose changes without silently altering stored preferences or context.

Setup and configuration

PLwC Personal targets Windows 11 x64. Python is required for the gateway, and Node.js is additionally required for the Chat Bridge. Docker Desktop configured for Linux containers, together with WSL 2, is required for document and sandbox operations. The Windows installer checks prerequisites and requests confirmation before downloading optional runtimes.

When Docker runtimes are unavailable, profile and workspace functions remain available in Safe Mode. Document and sandbox operations are disabled in that state. The official release includes installer checksums, a signed release manifest, runtime lock information, and runtime image digests for verification.

Tools and capabilities

The eight MCP tools provide these functional areas:

• Report gateway status and describe available behavior.
• Read or update local profiles and inspect reflection data.
• Plan and apply governed profile changes.
• Perform bounded workspace file and directory operations.
• Run Python or Node.js code in isolated containers.
• Process supported office, PDF, and ZIP documents.

Limitations and notes

PLwC Personal is a local gateway and tool set, not an AI model or client. You must supply a compatible assistant and model. Docker-dependent features are unavailable without the required runtimes, and the sandbox environments have no network access.

The release notes document a status display inconsistency: successful Python and Node.js executions may still report sandbox_ready: false, although the agreed execution tests completed successfully. The Windows installer is not Authenticode-signed; its release manifest has an independently verified RSA signature.

PLwC-owned components use the PolyForm Internal Use License 1.0.0, with stated permissions for personal use and official redistribution. Modified distribution and commercial offerings require a separate agreement. Third-party components retain their own licenses and rights.