Skip to content

README.md

Latest commit

 

History

History
194 lines (142 loc) · 9.58 KB

README.md

File metadata and controls

194 lines (142 loc) · 9.58 KB

cc-empire

The command center for a complete Claude Code workflow — hooks, rules, methodology, and the connective tissue between every tool in the ecosystem.

Built through 500+ real-world AI collaboration sessions by a non-developer. Every hook, every rule, every framework earned its place by solving an actual problem.

The Ecosystem

                              ┌─────────────────────┐
                              │      cc-empire        │
                              │  hooks + rules +     │
                              │  methodology + glue  │
                              └──────────┬──────────┘
                                         │
     ┌────────────┬───────────┬──────────┼──────────┬────────────┬────────────┐
     │            │           │          │          │            │            │
┌────┴────┐ ┌────┴────┐ ┌────┴───┐ ┌────┴────┐ ┌───┴───┐ ┌─────┴─────┐ ┌────┴────┐
│ Memory  │ │ Content │ │Pipeline│ │ Remote  │ │ Multi │ │  Clone    │ │Workflow │
│         │ │         │ │        │ │ Control │ │Session│ │           │ │Orchestr.│
│recallnest│ │alchemy+│ │alchemy │ │  TG +   │ │studio │ │ digital   │ │orchestr.│
│         │ │publisher│ │  new   │ │  WX     │ │       │ │  clone    │ │         │
└─────────┘ └─────────┘ └────────┘ └─────────┘ └───────┘ └───────────┘ └─────────┘

Modules

Module Stars Description Category
recallnest 8 Persistent memory via LanceDB. Hybrid retrieval, 40 MCP tools Core
content-alchemy 16 5-stage article pipeline with multi-perspective review Content
content-publisher 3 Image generation, layout, and WeChat publishing Content
workflow-orchestrator new Natural language pipeline: ideation, multi-role review, auto-routing Orchestration
telegram-ai-bridge 7 Run N parallel CC sessions from your phone Remote
wechat-ai-bridge 4 Full CC in WeChat, not an API wrapper Remote
claude-code-studio 2 Multi-session collaboration platform Orchestration
digital-clone-skill 1 Build digital clones from corpus data Content

All modules are optional. Install what you need.

Quick Start

git clone https://github.com/AliceLJY/cc-empire.git
cd cc-empire
bash install.sh

The installer will:

  1. Check prerequisites (Claude Code, git, etc.)
  2. Show available modules — pick what you need
  3. Install hooks into ~/.claude/hooks/
  4. Set up CLAUDE.md and workflow templates
  5. Clone selected module repos

What's Included

Hooks (12 scripts, 5 categories)

Production-grade hooks covering all 6 Claude Code lifecycle events.

Safety Guards

Hook Trigger What it does
bash-guard.sh PreToolUse:Bash Blocks rm, enforces tmux for dev servers, PR guards
edit-guard.sh PostToolUse:Edit README bilingual check, API signature change alerts
secret-guard.sh UserPromptSubmit Detects API keys/secrets before they leak

Workflow Automation

Hook Trigger What it does
auto-commit.sh Stop Reminds about uncommitted changes at session end
nag-reminder.sh PostToolUse ReAct drift detection — nudges observation notes
track-edit.sh PostToolUse:Edit Tracks which repos were touched per session

Lifecycle Management

Hook Trigger What it does
load-context.sh SessionStart Injects identity and project status on startup
pre-compact.sh PreCompact Saves active state before context compression
session-summary.sh Stop Auto-generates session index for later lookup

Quality Checks

Hook Trigger What it does
ts-check.sh PostToolUse:Edit Auto TypeScript syntax check after edits

Feedback Capture

Hook Trigger What it does
rating-capture.sh UserPromptSubmit Captures 1-10 ratings for quality tracking

All hooks support profile control (CC_HOOK_PROFILE=off to disable all) and individual disable (CC_DISABLED_HOOKS=hook1.sh,hook2.sh).

See docs/hook-lifecycle.md for the complete lifecycle diagram.

Rules & Frameworks

Distilled from 500+ sessions into reusable patterns. Located in rules/.

Methodology — How to work with AI

Rule What it solves
research-plan-implement Three-stage discipline. No skipping.
react-loop Act → Observe → Reflect → Record. Every step.
task-contracts Define completion criteria before starting
scope-drift-detection Catch when work drifts from the original goal
harness-engineering Shape AI behavior through rules, not hope
skill-evolution How skills improve through feedback loops

Thinking Frameworks — How to think clearly with AI

Framework What it does
anti-sycophancy Neutral prompts that get honest answers
bias-correction-matrix Systematic bias detection and correction
xy-problem-detection Catch when means and goals are misaligned
confidence-impact-filter Prioritize by confidence level and impact
completion-taxonomy DONE / DONE_WITH_CONCERNS / BLOCKED / NEEDS_CONTEXT

Borrowed Patterns — Proven techniques adapted for AI workflows

Pattern Origin
boil-the-lake Exhaustive enumeration before optimization
failure-modes-five-layers Systematic failure analysis
measurable-outcome Every skill needs a measurable success criterion
report-first-pattern Report findings before acting on them
two-stage-classification Rough sort, then fine-grade

Architecture — System design patterns

Doc Covers
memory-architecture How persistent memory is structured
confidence-tagging Tagging claims with confidence levels
reply-discipline Scan before replying, don't miss items

Templates

  • CLAUDE.md.example — Project contract template (execution rules, verification, reply discipline)
  • workflow.md — Research → Plan → Implement methodology with ReAct loop
  • memory/ — Memory index structure (MEMORY.md + identity + feedback templates)

Configuration

Environment Variables

Variable Default Purpose
CC_HOOK_PROFILE standard Set to off to disable all hooks
CC_DISABLED_HOOKS — Comma-separated list of hooks to skip
CC_MEMORY_DIR auto-detect Override project memory directory
CC_GITHUB_ORGS — Your GitHub org names (comma-separated), for PR guard
CC_NAG_THRESHOLD 5 Tool calls before ReAct reminder triggers

Tracked Repos

Edit ~/.claude/hooks/repos.conf to add your project directories:

TRACKED_REPOS=(
  "$HOME/Projects/my-app"
  "$HOME/Projects/my-lib"
)

Philosophy

This project was built by a non-developer through AI collaboration. The perspective is different from most CC workflow tools:

  • Cross-domain, not just coding — Writing pipelines, bot operations, remote control, not just TDD
  • Production-tested rules — Every rule exists because its absence caused a real problem
  • Memory as infrastructure — Persistent memory isn't optional; it's how the AI gets better over time
  • Honest methodology — Anti-sycophancy prompts, neutral verification, confidence tags
  • First-principles thinking — Patterns are cross-domain; a writing review framework becomes a code review framework

Architecture

See docs/architecture.md for how modules work together.

Contributing

Issues and PRs welcome. If you've built hooks or workflow rules that made your CC life better, share them.

License

MIT