
<!-- Copyright (c) 2026 OpenGuardrails.com Author: thomas-security <[email protected]> SPDX-License-Identifier: Apache-2.0 --> --- name: thomas-security description: | Use this skill when the user asks you to check, scan, audit, or red-team the security of an AI agent — either yourself, another agent installed on the same machine (OpenClaw, Claude Code, Cursor, Hermes), or a remote agent endpoint. Triggered by phrases like "give my agent a security check", "scan yourself
<!-- Copyright (c) 2026 OpenGuardrails.com Author: thomas-security <[email protected]> SPDX-License-Identifier: Apache-2.0 --> --- name: openguardrails description: | Use this skill when the user asks you to check, scan, test, or audit the security of an AI agent — either yourself, another agent installed on the same machine (e.g., OpenClaw, Hermes, Claude Code, Cursor), or a remote agent endpoint. Triggered by phrases like "give my agent a security check", "scan yoursel
<!-- Copyright (c) 2026 OpenGuardrails.com Author: thomas-security <[email protected]> SPDX-License-Identifier: Apache-2.0 --> --- name: thomas-security-contribute description: | Use this skill when the user asks you to add, draft, or submit a new security checkup, red-team test, or integration to the `thomas-security` repository. Triggered by phrases like "add a checkup for <CVE>", "write a red-team attack for <incident>", "contribute this advisory to thomas", "add an i