Workfront

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

Before installing: (1) review the full SKILL.md to confirm it does not instruct the agent to read arbitrary local files, shell history, or environment variables; (2) ask the publisher how Workfront authentication is performed — where are API tokens stored and who can access them (Membrane proxy vs skill-level requests); (3) confirm network endpoints the skill will call (should be Workfront API and Membrane only); (4) if you operate sensitive data, prefer only enabling this skill as user-invoked and test in a limited account first; (5) if anything in the full SKILL.md is unclear or requests unrelated system access, do not install until clarified.

Type: OpenClaw Skill Name: workfront Version: 1.0.1 The SKILL.md file contains an extremely large, repetitive, and malformed section of nested 'Data Extract View Pivot Value' strings that appears to be a context-stuffing or prompt-injection attempt designed to exhaust the agent's context window or manipulate its attention. Furthermore, the instructions require the agent to execute high-risk system commands, including a global package installation ('npm install -g @membranehq/cli') and interactive authentication flows. While these actions are framed as part of the 'Membrane' integration platform, the combination of malformed documentation and significant side effects on the host system is highly irregular.