← Back to Skills Marketplace
PPQ.AI Private Mode
by
mattius459
· GitHub ↗
· v1.0.1
· MIT-0
317
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install ppq-private-mode
Description
Install PPQ.AI Private Mode for end-to-end encrypted AI inference in OpenClaw
Usage Guidance
This skill does what it says (sets up a local PPQ proxy and configures OpenClaw) but requires you to install code from a GitHub repo and to store your PPQ API key in your OpenClaw configuration. Before installing: (1) verify the GitHub repository and its owner (PayPerQ) and inspect the code or releases (prefer a pinned release commit or checksum); (2) back up ~/.openclaw/openclaw.json; (3) consider creating a revocable or limited-scope PPQ key if possible; (4) prefer to run the install commands manually rather than allowing an agent to do them autonomously; (5) after installation, review the plugin's code and network behavior (e.g., confirm it only proxies to PPQ endpoints and doesn't exfiltrate keys). If you don't trust the repo or cannot review the code, do not install.
Capability Analysis
Type: OpenClaw Skill
Name: ppq-private-mode
Version: 1.0.1
The skill instructions direct the AI agent to perform high-risk system modifications, including installing a plugin from an external GitHub repository (https://github.com/PayPerQ/ppq-private-mode-proxy), modifying the core `~/.openclaw/openclaw.json` configuration file, and restarting a systemd service. While these actions are consistent with the stated purpose of installing the PPQ.AI Private Mode extension, the requirement for the agent to execute shell commands for remote code installation and system configuration changes represents a significant security risk and attack surface, as defined in the analysis criteria for risky capabilities.
Capability Assessment
Purpose & Capability
Name and description match the runtime instructions: the skill asks for a PPQ API key, installs a PPQ proxy plugin, updates OpenClaw's config to point at a local proxy, and restarts the gateway. The requested artifacts (API key, plugin install, config changes) are expected for enabling a local encrypted proxy workflow.
Instruction Scope
Instructions are narrowly scoped to installing a plugin, merging JSON into ~/.openclaw/openclaw.json, and restarting the OpenClaw gateway. These actions are in-scope for installing a provider plugin, but they touch persistent user configuration and a user systemd service. The doc does not instruct any broad system scans or access to unrelated files, but it also does not require verifying the plugin before installation.
Install Mechanism
There is no embedded install spec, but the instructions call out to `openclaw plugins install https://github.com/PayPerQ/ppq-private-mode-proxy`. Downloading and running code from a third‑party GitHub repository is common but inherently risky without integrity checks or a pinned release; the instructions include no verification (commit hash, checksum, or release tag).
Credentials
The only secret requested is the user's PPQ API key (sk-...), which is proportionate to a plugin that must authenticate to PPQ. However, the instructions require persisting that key in OpenClaw's plugin config, which is sensitive — the skill does not provide guidance on secure storage, rotation, or least-privilege keys.
Persistence & Privilege
The instructions modify a user config file (~/.openclaw/openclaw.json) and restart the user's openclaw-gateway.service. Those are reasonable for plugin installation but are high-impact operations. Because the platform allows autonomous invocation by default, an agent could perform these changes automatically. The skill does not request 'always: true', but the potential for automated modification of persistent config and service restart increases risk and warrants user confirmation and code review before running.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install ppq-private-mode - After installation, invoke the skill by name or use
/ppq-private-mode - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
PPQ Private Mode v1.0.1 Changelog
- Simplified API key setup instruction: always prompt user for their PPQ API key instead of auto-detecting existing keys.
- Revised plugin and provider configuration steps for clarity and conciseness.
- Streamlined model availability listing and switching instructions.
- Improved formatting and merged duplicate information in setup documentation.
- Troubleshooting section and some explanatory text removed for brevity.
v1.0.0
PPQ.AI Private Mode 1.0.0 – Initial release with end-to-end encryption for OpenClaw
- Queries are encrypted on the user's machine and only decrypted in a secure hardware enclave.
- Step-by-step setup instructions provided, including API key management, plugin installation, and configuration.
- Supports five private models and auto-routing with detailed switching instructions.
- Built-in troubleshooting steps for common issues, including API key, attestation, and port conflicts.
- No user data is visible to PPQ.AI or intermediaries; all inference is hardware-secured.
Metadata
Frequently Asked Questions
What is PPQ.AI Private Mode?
Install PPQ.AI Private Mode for end-to-end encrypted AI inference in OpenClaw. It is an AI Agent Skill for Claude Code / OpenClaw, with 317 downloads so far.
How do I install PPQ.AI Private Mode?
Run "/install ppq-private-mode" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is PPQ.AI Private Mode free?
Yes, PPQ.AI Private Mode is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does PPQ.AI Private Mode support?
PPQ.AI Private Mode is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created PPQ.AI Private Mode?
It is built and maintained by mattius459 (@mattius459); the current version is v1.0.1.
More Skills