Usercentrics

Usercentrics integration. Manage data, records, and automate workflows. Use when the user wants to interact with Usercentrics data.

This skill appears to be a Membrane-based Usercentrics connector but the package metadata omits the credentials/config it claims to need. Before installing or enabling it: (1) ask the publisher which credentials are required (e.g., MEMBRANE_API_KEY or USERCENTRICS_API_TOKEN) and why they are needed; (2) verify the skill's source repository and homepage (the registry lists getmembrane.com and a membranedev GitHub repo — confirm those are legitimate and contain the referenced skill code); (3) request example API calls or a minimal runbook showing what the agent will send/receive so you can ensure no unrelated secrets or local files are accessed; (4) avoid installing until required env vars and network endpoints are explicitly declared. If you must test it, do so in a restricted environment with least-privilege credentials and monitor outbound traffic and logs.

Type: OpenClaw Skill Name: usercentrics Version: 1.0.1 The SKILL.md file contains an extreme amount of repetitive keyword stuffing (hundreds of repeated verbs and nouns like 'Manage', 'Control', 'Direct'), which is a technique often used in prompt injection to manipulate an AI agent's context or bypass behavioral constraints. Additionally, the instructions require the agent to perform high-risk system operations, including a global NPM installation (npm install -g @membranehq/cli) and executing CLI commands that handle authentication. While these actions are framed as part of the legitimate Membrane platform, the highly irregular documentation structure and the promotion of global package installation without verification warrant a suspicious classification.