superclaude-plusplus
CLAUDE.md14 starsChanged 37 days agoMITView on GitHub
Always-resident context is limited to the four files above. Everything else loads on demand from optional/ or as skills. Rules restate neither harness guarantees nor model defaults — see the preamble in RULES.md. Skill names/descriptions and available agent types are auto-injected by the harness every session — they are not listed again here. Agents enforce model, tools, maxTurns, effort via AGENT.md frontmatter. - Auto-Invoke: UserPromptSubmit hook (scripts/skill-matcher.py + .claude/skill-rules.json) — mechanical prompt/file matching - Manual: /skill-name | Proactive: suggestion with…
In this repository44 other files it gives its agents.
- adaptSkill
- animateSkill
- auditSkill
- bolderSkill
- brainstormSkill
- build-fixSkill
- business-panelSkill
- checkpointSkill
- clarifySkill
- colorizeSkill
- composition-patternsSkill
- confidence-checkSkill
- config-doctorSkill
- critiqueSkill
- delightSkill
- design-auditSkill
- devils-advocateSkill
- distillSkill
- eval-harnessSkill
- feature-plannerSkill
- fix-prSkill
- frontend-designSkill
- gap-analysisSkill
- grill-with-docsSkill
- hardenSkill
- impeccableSkill
- layoutSkill
- learnSkill
- optimizeSkill
- overdriveSkill
- polishSkill
- pytest-runnerSkill
- python-best-practicesSkill
- quieterSkill
- react-best-practicesSkill
- sc-helpSkill
- security-auditSkill
- shapeSkill
- tddSkill
- typesetSkill
- ui-ux-pro-maxSkill
- uv-packageSkill
- verifySkill
- web-design-guidelinesSkill
The file
We split this CLAUDE.md into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Always-resident context is limited to the four files above.
≈ “Common” is an estimate: at least 1 in 5 of the 26,980 CLAUDE.md 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
SuperClaude++ v3.4
Language
Response language is the harness language setting (config/settings.json:
한국어), not a rule here. What the setting does not cover:
- Code comments/variables: English
- Technical terms: English when common (WebSocket, API, etc.)
- Korean fluency (조사/어미 보존, 번역투 방지):
fluent-koreanoutput style (snflkd/fluent-korean 플러그인,config/settings.json에서 선언 — provenance: NOTICE.md)
Core Framework
@RULES.md @PRINCIPLES.md @MODES.md @CONVENTIONS.md
Always-resident context is limited to the four files above. Everything else
loads on demand from optional/ or as skills. Rules restate neither harness
guarantees nor model defaults — see the preamble in RULES.md.
Skill & Agent System
Skill names/descriptions and available agent types are auto-injected by the
harness every session — they are not listed again here. Agents enforce
model, tools, maxTurns, effort via AGENT.md frontmatter.
- Auto-Invoke:
UserPromptSubmithook (scripts/skill-matcher.py+.claude/skill-rules.json) — mechanical prompt/file matching - Manual:
/skill-name| Proactive: suggestion with confirmation - Suggestion intensity:
--suggest-all(default) |--suggest-minimal|--suggest-off - Third-party skill provenance:
NOTICE.md
Workflow Integration
- Step 0: Difficulty Assessment (Simple/Medium/Complex) → protocol branching
- Pre-Implementation:
/confidence-check→ ≥90% proceed (Medium+) - Goal Lock (autonomous mode): Strong success criteria + multi-turn work →
/goal "<verifiable condition>"(seeoptional/GOAL_PATTERNS.md) - Planning:
/feature-planner - Design:
DESIGN.md(if exists) →/ui-ux-pro-max→/frontend-design→/web-design-guidelines - Review: Two-Stage Review (Simple: Stage 1 only | Medium: Stage 1+2 | Complex: + Cascade Impact)
- Verification:
/verify→/audit(project rules) - Post-Implementation:
/learn→/goal clear(if a goal was set)
DESIGN.md (Visual Design System)
AI 에이전트가 읽는 디자인 시스템 문서 (Google Stitch format).
| File | Who reads it | What it defines |
AGENTS.md | Coding agents | How to build the project |
DESIGN.md | Design agents | How the project should look and feel |
Usage: npx getdesign@latest add {brand} (66 brands) | /ui-ux-pro-max --design-system --persist | Template: templates/visual-design.template.md
On-Demand References (optional/, load when needed)
FLAGS.md— behavioral flag definitions (--think,--uc,--delegate, ...)CONTEXTS.md— DEV/REVIEW/RESEARCH/PLANNING context modesMCP_SERVERS.md— MCP server selection matrix; per-server guides inMCP_*.mdGOAL_PATTERNS.md—/goalcondition patterns, anti-patterns,/loopvs/goalOVERENGINEERING_TRAPS.md— Build Ladder application rules + rung 3 native-feature catalogREASONING_TEMPLATES.md— structured reasoning (debugging, architecture, performance)WORKER_TEMPLATES.md— worker agent prompt templatesPROTOCOLS.md— learning protocol, built-in memory layoutPROJECT_RULES.md— Dockerfile/CI patterns, security checklist
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.