ADHD skill
A parallel divergent ideation skill for agents: fans out under cognitive frames, prunes traps and deepens the best branches
- 4,538 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Spawns several parallel model calls, which consume tokens and reach the network
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/adhdDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Spawns several parallel model calls, which consume tokens and reach the network
This is third-party code. Review the repository files before installing.
What it does
The skill fights the model's habit of latching onto its first answer. It spawns several isolated reasoning branches under deliberately skewed cognitive frames, for example regulator, biology, speedrunner, a ten-year-old or a zero budget, and during divergence the branches do not see each other, so there is no shared anchor. A separate critic pass then scores the ideas, clusters them, prunes traps and deepens the survivors. It suits open-ended tasks: architecture decisions, fuzzy debugging, naming, API surface design, strategy and any prompt of the shape give me a few ways to. Before running, the skill goes through a pre-flight gate and refuses when a direct answer is enough, because a run is expensive. It is available as a skill, a CLI and a library, and installs into many agents with one command.
Who it is for. For developers and teams who need to explore options on open-ended design and strategy questions.
Good fit when
- An open-ended decision on architecture, API design or strategy
- Fuzzy debugging with no known root cause
- You need several options with the weak ones pruned
Not a fit when
- There is one canonical answer, or it is a syntax or lookup question
- A bug with a known root cause
- You are asking for a quick or standard answer
Example request
Run ADHD and give me a few ways to shard the queue under bursty loadLimitations
A run is expensive: about ten agent calls, thirty to ninety seconds and noticeably more tokens than a single answer. Node 18 or newer is required. The skill runs on top of your agent's model and relies on the Claude and Codex Agent SDK, so cost and availability depend on your access to those models. Output quality depends on the model.
How to disable. Remove the skill folder from your agent's skills directory or uninstall the plugin via /plugin. The standalone CLI is removed with npm uninstall -g adhd-agent.
FAQ
How is this different from chain-of-thought and tree-of-thought?
The divergence branches are isolated and share no context, so the first answer does not pull the rest along. Evaluation happens in a separate critic pass after generation.
Is it expensive?
Yes. That is why the skill has a pre-flight gate that skips a run when a direct answer is enough.
Discussion
No questions or reviews yet.
Related
SkillsA skill that convenes a council of thinker personas and runs a structured deliberation on a hard decision, ending with a verdict
A skill plus hooks that keep a coding agent on task and cut speculative defenses, unread checksums and scope creep
Skill that rewrites a task into a bolder brief: the goal beyond the goal, a top specialist approach, methods from other fields.
A skill with 25 commands to audit and improve the prompts, tools and agent designs of a project in Cursor, Claude Code and Codex
Agents Best Practices
agents-best-practices
A provider-neutral skill for designing agent systems: loop architecture, tools, permissions, memory and observability
A skill that links two to four open Codex CLI windows: one splits the work while the others execute, cross-check and deliver.
README in short
The README frames ADHD as an architectural fix for premature convergence in a model's reasoning. Linear chain-of-thought anchors on its first statement, tree-of-thought widens the search but walks a shared context, so the anchor persists. ADHD spawns isolated branches under different cognitive frames with no shared context during divergence, then a separate pass scores, clusters and prunes traps. It walks through one problem with an independent judge scoring breadth, novelty and trap detection. Installation is a single npx skills add command with agent auto-detection, and a CLI and library are also available. MIT licensed.
README badge
Original SKILL.md
---
name: adhd
description: Parallel divergent ideation for coding agents. Spawns N isolated branches under different cognitive frames (regulator, biology, speedrunner, 10-year-old, $0 budget), scores, clusters, prunes traps, and deepens top survivors. Use on /adhd, "ADHD mode", brainstorm/ideate intents, or open-ended design, architecture, naming, API/SDK surface, and fuzzy-debugging decisions. Skip for syntax, lookups, bugs with known root cause, or closed phrasing ("quick", "standard", "canonical", "textbook"). Full pre-flight gate is in the skill body.
license: MIT
---
# ADHD
Stop picking the textbook answer. The first three answers the model would
give are the answers a senior engineer would give in thirty seconds.
Correct. Forgettable. The interesting answers live past number three, in
the awkward middle nobody walks into. This skill makes the model walk
there.
## Pre-flight (run before Phase 1)
This skill is expensive. About 10 Agent calls, 30 to 90 seconds wall clock,
5 to 10x a single answer. Do not pay that cost when a direct answer is
better. Run this gate before Phase 1.
**Step 1. Explicit invocation check.**
If the user typed `/adhd` or explicitly asked for ADHD mode, "use the
adhd skill", or "run ADHD on this", **SKIP the rest of this section and go
straight to Phase 1**. The user opted in. Do not second-guess.