Generates divergent coding ideas through parallel isolated cognitive frames with scoring and clustering.
npx skills add https://github.com/visdomtech/skills --skill adhd-visdomtech
Please help me install this Agent Skill. Skill: adhd Source: https://github.com/visdomtech/skills/tree/main/adhd Command: npx skills add https://github.com/visdomtech/skills --skill adhd-visdomtech
High-intent search queries and answers about installing and using this skill.
Invoke the skill with /adhd followed by your problem, or ask for ADHD mode explicitly. It spawns 5 parallel isolated branches under different cognitive frames, then scores, clusters, and deepens the top 3 ideas into actionable sketches.
Divergent ideation generates many distinct candidate approaches before evaluating any of them. This skill separates generation from criticism into two strict phases, using frames like regulator, biology, and speedrunner to push past the first three obvious answers.
Skip it for syntax questions, lookups, bugs with a known root cause, or requests phrased with words like quick, standard, canonical, or textbook. The loop costs roughly 10 agent calls, so it only pays off on open-ended high-stakes decisions.
Branches that see each other's output anchor on the same ideas and collapse into a wider single thought. Isolated parallel agent calls give each branch a fresh context, which preserves genuine diversity across the candidate pool.
Yes, there is a Node/TypeScript implementation installable via npm install -g adhd-agent, runnable as adhd "your problem here". It performs the same loop with structured JSON parsing and score weighting for batch use.
Nandansai08Converts vague research topics into pinned, answerable questions with decision links and timeboxes.
mnfst-aiIdentify load-bearing assumptions in product and design documents and rank them by cost-if-wrong.
Nir-BhayEvaluate code review feedback against codebase constraints and the YAGNI principle.
microsoftCoach UX practitioners through problem framing, design critiques, and stakeholder advocacy conversations.
