/toxic-review
Use for a strict, skeptical, risk-focused review of a named work product without attacking people or inventing criticism.
$ npx -y skills add dkstm95/hope --skill toxic-review --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.
- You can call itInvoke it directly when you want it.
- Slash command
/toxic-review
Context preview
The summary Claude sees to decide when to auto-load this skill.
Use for a strict, skeptical, risk-focused review of a named work product without attacking people or inventing criticism.
Hope helps people work with AI while staying able to see, understand, and control the work.
Other skills on hope.
- /align
Use when someone asks for a pre-implementation shared-understanding check or wants requirements, scope, design, or expected behavior clarified before coding.
Open skill - /diff
Use to explain a GitHub pull request as an evidence-linked review, including a pending Hope Diff confirmation.
Open skill - /polish
Use for one bounded cleanup or refactor of a named completed work product while preserving behavior and meaning.
Open skill - /settings
Use to show, set, or reset Hope's language or theme.
Open skill - /sweep
Use to inspect a codebase for broad maintenance, show a bounded plan, and apply only exact approved behavior-preserving work.
Open skill - /write
Use for a standalone language-only draft, edit, or review that must preserve meaning, facts, uncertainty, and voice.
Open skill