agentleFS
rahmanef63rahmanef63/si-coder-agent/skills/sc/SKILL.md

sc

Skill13 starsChanged 2 months agoMITView on GitHub

Main SI-Coder entry point for non-technical users. Turn a plain-language idea or existing web app into a working product, improve frontend quality across UI/UX/DX/AX, and publish it. Route automatically while keeping technical details optional.

The file

We split this skill into the rules it gives an agent

7 rules. About 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write the skill instructions and documentation in English, but reply in the user's language unless the user asks for another language.

    line 24
  2. Distinctive:

    Do not ask the user to choose a SI-Coder sub-skill.

    line 28
  3. Distinctive:

    If sibling SI-Coder skills are installed, route to them normally.

    line 49
  4. Very specific:

    If they are not installed, use the bundled files under references/si-coder/ inside the package.

    line 50
  5. Very specific:

    Do not tell a web user to install a local CLI merely because sibling skills are absent.

    line 51
  6. Distinctive:

    Ask one question at a time.

    line 62
  7. Very specific:

    After a meaningful completion, show exactly one [rekomendasi] block with the highest-value next step, why it helps, what is required, and a …

    line 79

≈ “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.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

sc — the main SI-Coder entry point

sc is the canonical SI-Coder skill identity. Invocation syntax is chosen by the host product, not by the .skill package itself.

Examples:

  • ChatGPT Web: @sc Create a booking app for my salon with customer and admin access, then put it on my domain. or ask naturally and let ChatGPT select sc automatically.
  • Claude Code: /sc Create a booking app for my salon with customer and admin access, then put it on my domain.
  • Frontend quality on slash-capable hosts: /sc-fe --workbench improve the desktop shell or /sc-fe --apple refine settings.

Language

Write the skill instructions and documentation in English, but reply in the user's language unless the user asks for another language.

Intent routing

Do not ask the user to choose a SI-Coder sub-skill.

  • New or vague product idea → follow the sc-build skill.
  • Existing app that needs frontend/UI/UX improvement or any substantial frontend work → follow sc-fe.
  • Focused visual interface request → sc-ui through sc-fe unless the user explicitly invokes sc-ui.
  • Focused usability/accessibility/interaction request → sc-ux through sc-fe unless explicitly invoked.
  • Focused frontend developer-experience request → sc-dx through sc-fe unless explicitly invoked.
  • Focused Agent Experience/tool ergonomics request → sc-ax through sc-fe unless explicitly invoked.
  • Existing app that needs to go live, change hosting, or attach a domain → follow the sc-all skill; sc-all delegates user-facing frontend quality to sc-fe before final verification when applicable.
  • Account/permission/API access task → follow the sc-provider skill.
  • Installation into another agent runtime → follow the sc-install skill.
  • Explicit advanced provider operation → use the matching sc-* provider skill only when that skill is active/implemented. Unfinished/stub provider work stays in sc/sc-provider; explain the limitation instead of routing to a dead-end skill.

Forward frontend flags such as --apple, --workbench, --profile <name>, --save-profile <name>, --density, --motion, --platform, --audit, and --strict to sc-fe rather than interpreting them as deployment flags.

The route is internal. Do not narrate the skill handoff unless it helps recover from a problem.

Standalone package mode

This skill may be installed either as part of the full SI-Coder repository/plugin or as the one-file sc.skill web package.

  • If sibling SI-Coder skills are installed, route to them normally.
  • If they are not installed, use the bundled files under references/si-coder/ inside the package.
  • Do not tell a web user to install a local CLI merely because sibling skills are absent. Hosted web execution should use the tools/apps/connectors available on that surface.

Surface invocation is not universal: Claude Code can invoke this as /sc; ChatGPT Web currently documents automatic activation or explicit @sc selection; Codex uses its own current skill-selection/invocation UX. A .skill file packages the skill but does not register a custom ChatGPT Web slash command.

Non-technical default

Lead with outcomes. Hide stack, hosting vendor, database vendor, repository mechanics, DNS, environment variables, containers, deploy keys, provider routing, and internal quality-axis routing unless:

  1. the user explicitly asks for technical details, or
  2. one technical fact is necessary for a user action or error recovery.

Ask one question at a time. Prefer a useful default over asking the user to make a technical choice. Do not ask a question that repository/tool state can answer.

For existing products, preserve coherent UI/UX design DNA by default. Do not replace it with a generic preset unless the user asks for a redesign or explicit preset.

Project initiation through CI/CD

For launch, domain migration, auth/email setup or CI/CD, read the delivery workflow before following the selected sub-skill. It covers the complete requested outcome, including backend release, OAuth/legal pages, transactional email, exact revision verification and handoff. Apply only the services/features the project needs. Resume verified work and preserve existing architecture and the user's prior authorization.

Completion contract

A "done" result means the requested user-facing outcome works. For a published app this normally includes the app, data path, public URL/domain, HTTPS, frontend quality verification when applicable, and a basic functional verification.

After a meaningful completion, show exactly one [rekomendasi] block with the highest-value next step, why it helps, what is required, and a simple opt-in.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.