Security-first behavioral guidelines for cautious agent operation. Use this skill for ALL operations involving external resources, installations, credentials, or actions with external effects. Triggers on - any URL/link interaction, package installations, API key handling, sending emails/messages, social media posts, financial transactions, or any action that could expose data or have irreversible effects.
Security Analysis
high confidenceAn instruction-only 'zero trust' checklist whose requirements and behavior are consistent with its stated purpose and which does not request credentials, installs, or elevated privileges.
The skill's name and description (a conservative security checklist) match the SKILL.md instructions. It requests no env vars, binaries, or installs that would be unrelated to a 'zero trust' policy.
The runtime instructions are appropriately restrictive and focused on verifying external interactions. Minor issues: the guidance hardcodes asking approval from a person named 'Pat' (unclear who that maps to in your workflow), and some items are prescriptive/opinionated (e.g., always store credentials under ~/.config/). The instructions could produce many human approval prompts, which may materially slow or block agent operations.
No install spec and no code files — lowest-risk form. Nothing is downloaded or written to disk by the skill itself.
The skill requests no environment variables, credentials, or config paths beyond general storage recommendations. There is no disproportionate access requested.
The skill is not force-included (always:false) and is user-invocable. It does not request system-wide changes or modify other skills' configs.
Guidance
This skill is essentially a conservative checklist and is internally coherent and low-risk because it has no install or credential requirements. Before enabling it: (1) confirm who the skill means by the human approver ('Pat') and how human approvals will be presented to you, (2) be aware it will demand many explicit approvals and could slow automated workflows, (3) review/personalize prescriptive items (credential storage path, what counts as 'trusted' search engines), and (4) test it in a controlled environment to ensure its prompts and STOP/ASK guidance integrate with your agent's UI/workflow. If you need automatic operations, this skill's strictness may be impractical without human-in-the-loop tooling.
Latest Release
v1.0.0
**Initial release: Establishes security-first protocols for all high-risk operations.** - Introduces zero-trust guidelines for agent operations involving external resources, installations, or credentials. - Requires explicit human approval for sensitive actions such as sending emails, installing packages, or clicking unknown links. - Outlines STOP → THINK → VERIFY → ASK → ACT → LOG flow for all external actions. - Defines strict credentials handling: never log or expose, always store securely. - Provides clear red flags to identify risky operations and immediate STOP criteria.
Popular Skills
Published by @doonot on ClawHub
