From deep-think
Use when the user conversationally asks to add review checklist items, custom reasoning strategies, or modify their deep-think review process. Triggers on phrases like "add a review step for...", "create a strategy for...", "I want to always check for... during review", "add a custom thinking framework".
How this skill is triggered — by the user, by Claude, or both
Slash command
/deep-think:manage-practicesThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
The user wants to add, remove, or view review practices and custom strategies.
The user wants to add, remove, or view review practices and custom strategies.
Practices map to .deep-think.json in the current project directory:
practices.reviewChecklist — items checked during reflectionstrategies.custom — custom reasoning strategies with steps and checkpoints.deep-think.json from the current project directory.
practices.reviewChecklist array.name — short identifier (kebab-case)description — when to use this strategysteps — ordered list of reasoning stepscheckpoints — things to verify at each stage.deep-think.json back to disk.npx claudepluginhub bis-code/mcp-deep-think --plugin deep-thinkGuides Claude through a 10-principle structured thinking loop for non-trivial problems: architectural decisions, post-mortems, ambiguous requests, audits, multi-stakeholder tradeoffs.
Manually triggers structural and/or cross-model adversarial review on deep-work Phase documents (brainstorm/research/plan). Supports phase targeting and selective review modes.