new-surface
Skill14 starsChanged 36 days agoApache-2.0View on GitHub
Scaffold a surface standard and portable specialist adapter triad, then register the surface in the hub.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- steepy-it/steepy-formulatheir own version, first seen Oct 25, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- steepy-it/steepy-formula★ 5same name
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 repository14 other files it gives its agents.
- AGENTS.mdAGENTS.md
- templates/AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- steepy-apex-bootstrapSkill
- steepy-apex-bootstrapSkill
- brainstormSkill
- checkSkill
- discoverySkill
- implementSkill
- inceptionSkill
- initSkill
- loop-engineerSkill
- planSkill
- reviewSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 2 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Add one governed surface through the common specialist planner.
- Very specific: line 18
Read .apex/_INDEX.md and gather the missing confirmed inputs:
- Distinctive: line 24
Run the surface scaffolder from this skill's engine root:
- Very specific: line 55
Do not reproduce planner provenance markers or hand-render an adapter in this skill.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
new-surface
Add one governed surface through the common specialist planner.
Engine root: this skill's base directory is
<engine-root>/skills/new-surface/; engine scripts live two levels up, at<engine-root>/scripts/. Resolve them relative to the base directory your harness reports for this skill.
Procedure
- Read
.apex/_INDEX.mdand gather the missing confirmed inputs:- surface
name, a lowercase hyphenated identifier; - repository-relative surface
path; - semantic specialist
agent, defaulting to<name>-agentonly when the user accepts it; - surface
testCmd.
- surface
Run the surface scaffolder from this skill's engine root:
node <engine-root>/scripts/new-surface.mjs \ --name <name> \ --path <path> \ --agent <agent> \ --test "<testCmd>" \ --hub <repo-root>new-surface.mjsfirst reports which explicit mode it used:- active-v1 — it derives the real Project identity from the canonical managed
AGENTS.md, adds or confirms the surface in that full Project model, and routes the change through the public planner inproject-scaffold.mjs. It updates the managed root and routing row together with the create-only standard and generated specialist triad. If the planner reports a conflict, make the user choose only from its emitted ID and choices, then repeat the command with one--resolution <id=choice>per choice. Do not infer a replacement or add an unoffered choice. - preparatory-unbound — when no managed Project identity exists yet, it explicitly limits itself to the common specialist planner. It plans all four surface artifacts before the first write and creates the standard plus specialist triad without pretending to update a Project identity that does not exist.
In both modes the surface result is:
.apex/standards/<name>.md;.claude/agents/<agent>.md;.codex/agents/<agent>.toml;.opencode/agents/<agent>.md.
Do not reproduce planner provenance markers or hand-render an adapter in this skill. The standard remains create-only. If the surface is a repair of an incomplete scaffold, add
--repair; repair keeps current artifacts and creates only absent members of the standard-plus-triad result. Any adapter that is not the current canonical rendering, an untouched earlier one included, is acustomizedconflict: repair stops, and only an explicitreplace(through--resolution <id>=replacein an active hub) writes the current rendering.- active-v1 — it derives the real Project identity from the canonical managed
- In active-v1 mode, do not edit generated or managed artifacts after the command: the public
workflow has already made the root, routing row, standard, and triad agree. In
preparatory-unbound mode only, append the emitted routing row to
.apex/_INDEX.mdif that surface has no row. Never rewrite the routing table or change another row. Verify:
node <engine-root>/scripts/validate-hub.mjs .Fix coherence errors before reporting completion. Then rerun with
--repair; a complete result must report every standard/triad member preserved and leave bytes, modes, and mtimes unchanged. Report each member of the standard-plus-triad result as created, updated, or preserved, and report every other file the command lists as created or updated.
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.

