next-step
Skill14 starsChanged 5 days agoMITView on GitHub
Use at every point of a conversation where one of this collection's skills would serve the person better than going on as is, and they have not named it: an idea still vague or choices left open (offer the interview), decisions worth keeping (a brainstorm board), agreement on something to build (a plan, or a swarm with its protocol filled in), a question to decide or a text to write from several viewpoints (a brainswarm), a run that just closed (its follow-up). Produces one offer in one line; never runs the skill unasked. Not for explaining how the collection works, mid-work or to a new person (help). Not for running any of those skills.
12 other projects use a version of this.
Adopted by
Most recently taken up by
- dotslashsimran/simcrafttheir own version, first seen Nov 8, 2026
- scoobydrew83/skillstheir own version, first seen Nov 3, 2026
- spongetimblin/claude-skill-librarytheir own version, first seen Oct 25, 2026
- Ovich/swarmstormword for word, first seen Oct 25, 2026
- SHYXIN/skillstheir own version, first seen Oct 11, 2026
- sd0xdev/sd0x-dev-flowtheir own version, first seen Oct 8, 2026
Other versions12 projects use a version of this.
Word for word · 1
- Ovich/swarmstorm★ 14100% same
Edited · 0
None found.
Wrote their own version · 11
- bitwize-music-studio/claude-ai-music-skills★ 589same name
- sd0xdev/sd0x-dev-flow★ 192same name
- sd0xdev/sd0x-harness★ 192same name
- dotslashsimran/simcraft★ 69same name
- SHYXIN/skills★ 6same name
- jordnlvr/MendixToolkit★ 2same name
- zahardev/aicontext★ 2same name
- jwd83/skills★ 1same name
- Life696969/next-step★ 1same name
- spongetimblin/claude-skill-library★ 1same name
1 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository34 other files it gives its agents.
- AGENTS.mdAGENTS.md
- brainstormSkill
- brainswarmSkill
- interviewSkill
- whiteboard-defenseSkill
- complexity-reworkSkill
- deepen-moduleSkill
- dependency-graphSkill
- simplify-reworkSkill
- simplifySkill
- write-diagramsSkill
- aiviewSkill
- swarm-boardSkill
- execute-planSkill
- long-horizonSkill
- orchestrate-planSkill
- guide-complexitySkill
- guide-createSkill
- guide-deepen-moduleSkill
- helpSkill
- orchestrateSkill
- cold-readSkill
- write-planSkill
- roadmapSkill
- technical-writingSkill
- write-conventionsSkill
- pr-reviewSkill
- recurring-review-findingsSkill
- swarmSkill
- design-prototypeSkill
- point-atSkill
- validate-deliverySkill
- verification-skill-createSkill
- create-workbenchSkill
The file
Files
Next step
The person should never have to know a skill's name to get its help. You notice the moment one fits and offer it in one line, with why and what it would produce; the person says yes, no, or something else. A yes loads that skill; it takes over from there.
The moments
| You notice | You offer | It produces |
| an idea still vague, several choices open, the person hesitating or contradicting an earlier answer | the interview skill: "Want me to interview you on this, one question at a time?" | the decisions settled one by one, a recommendation on each |
| decisions taken in chat that someone will need later | the brainstorm skill: "Keep this as a board?" | a board in aiview: the decisions, diagrams, what was looked into |
| agreement on something to build, larger than one sitting | the write-plan skill: "Write it as a plan?" | the plan, its phases and seats, run later on the board |
| agreement on work that splits into parts or must be seen working, no plan needed | the swarm skill, with the protocol filled in: "Run it as build-verify: 3 builders, verifiers sized from the steps (2 here) each on its own instance, until ready (cap 5), effort low. Adjust or go?" | the run on the board, the merged branch |
| a question to decide, options to explore, or a text to write, better from several viewpoints | the brainswarm skill, its protocol named: debate, brainstorm, editorial | the run's record in aiview |
| a screen to build or change | the design-prototype skill | a prototype in the project's design language |
| work on one subject going on past this session, with findings judged as it goes and no plan holding them | the long-horizon skill: "Keep this as a long-horizon document, so any session can resume it?" | the document in aiview: a summary and the elements, each with its status |
| a project's conventions just decided, or about to be | the write-conventions skill | the rule in the conventions file |
At a run's close, its follow-up: a brainstorm or a debate settled → a plan; a plan approved →
its run, with build-verify when the result must be walked (with a prototype phase first when it changes a screen, an architecture phase when its modules must be defined); a run whose setup will recur → keep it
as the project's protocol-<name> skill (the orchestrate skill's protocols); a delivery to go
over together → the validate-delivery skill; a run that changed code, its verdict ready → the
simplify skill on its range: "Simplify this run's range <base>..<head>, parity verified after?
Its builders still know what they replaced; later it costs a fresh exploration."; the person asking
where the dead code is with no range → the simplify-rework skill.
How you offer
- One line, one recommendation, the moment it fits, not before: what you would run, why now, what comes out. A second option only when two fit equally, named in the same line.
- A filled-in proposal beats a question: a swarm offer names its protocol, seats, loop and effort, so the person adjusts something concrete.
- Once per subject. A no ends it for that subject; never ask again unless the person comes back to it.
- Never while the person is mid-thought: at a pause, after an answer, at a decision.
- The person's own request always wins: they named a skill, you run that one.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.