Agent skill · research science · aradotso
claude-code-system-prompts-research
claude-code-system-prompts-research — an agent skill for coding agents.
Why this skill is useful
Adds extensive documentation and insights into Claude Code's internal architecture, including multi-agent orchestration patterns and security classifiers that the AI wouldn't generate on its own.
What it needs
About 7k tokens when loaded. Last updated 2026-07-09. 70 stars on the source repository.
What this skill does
claude-code-system-prompts/ README.md prompts/ 01mainsystemprompt.md # Master assembled prompt 02simplemode.md # CLAUDECODESIMPLE minimal prompt 03defaultagentprompt.md # Base inherited by all sub-agents 04cyberriskinstruction.md # Security allow/deny boundaries 05coordinatorsystemprompt.md # Multi-worker orchestrator 06teammatepromptaddendum.md # Swarm/team communication protocol 07verificationagent.md # Adversarial testing specialist 08exploreagent.md # Read-only codebase explorer 09agentcreationarchitect.md # Designs new agent configs 10statuslinesetupagent.md # Terminal status line setup 11permissionexplainer.md # Tool risk explanations 12yoloautomodeclassifier.md # 2-stage security classifier 13toolprompts.md # All 30+ tool descriptions 14toolusesummary.md # Git-commit-style tool labels 15sessionsearch.md # Semantic session search 16memoryselection.md # Memory file relevance selector 17automodecritique.md # Reviews classifier rules 18proactivemode.md # Autonomous tick-based agent 19simplifyskill.md # 3-agent parallel code review 20sessiontitle.md # Session title generator 21compactservice.md # Conversation summarization 22awaysummary.md # Idle session recap 23chromebrowserautomation.md # Browser extension integration 24memoryinstruction.md # CLAUDE.md @include directives 25skillify.md # Skill creation interview 26stuckskill.md # Frozen session diagnostics 27rememberskill.md # Promotes memory to CLAUDE.md 28updateconfigskill.md # settings.json/hooks manager 29agentsummary.md # Sub-agent progress updates 30promptsuggestion.md # Predicts next user command ┌─────────────────────────────────────────┐ │ STATIC PREFIX (cached globally) │ │ - Identity + Cyber Risk │ │ - Permission modes, hooks, reminders │ │ - Code style, security, reversibility │ │ - Tool preferences, parallel calls │ │ - Tone, output efficiency patterns │ ├─────────────────────────────────────────┤ │ CACHE BOUNDARY │ ├─────────────────────────────────────────┤ │ DYNAMIC SUFFIX (session-specific) │ │ …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aradotso/claude-code-system-prompts-research