> **Auto-Generated**: This file is synchronized from `instructions/`. Do not edit manually.
# Cursor Rules & Instructions
> **Auto-Generated**: This file is synchronized from `instructions/`. Do not edit manually.
<!-- BEGIN_GENERATED_INSTRUCTIONS -->
# AI Agent Operational Guide & Governance
> **Context**: This repository (`ai-dev-tasks`) is a shared **framework/submodule** used by other projects.
> Agents operating here are **Framework Contributors**, not Consumer Users.
## 1. 🤖 Core Mandates & Governance
### **Claudia is Boss**
- **Authority**: If `claudia-claude-code-agent` is running, respect her "STOP" commands immediately.
- **Coordination**: Claudia coordinates multi-agent workflows. Defer to her judgment on conflicts.
- **Monitoring**: Ensure Claudia is monitoring the session if possible.
### **Agent-First Workflow**
- **Consult First**: Before starting any significant task, ask: "Which agent can help with this?"
- **Delegation**:
- **Script Logic**: Use `code-reviewer`
- **Auth/Security**: Use `security-reviewer`
- **Architecture**: Use `cto-agent`
- **Docs/Specs**: Use `spec-writer`
- **Requirements**: Use `business-analyst`
- **Reference**: See `AGENTS.md` or `docs/reference/AGENT_REFERENCE.md` for full capabilities.
### **95% Certainty Rule**
- Do not proceed with an action unless you are >95% certain.
- If uncertain, ask clarifying questions or consult a specialized agent.
---
## 2. 📚 Documentation Retrieval Policy
### **Context7 Integration**
- **Mandate**: When using external libraries, APIs, or frameworks, **ALWAYS** retrieve up-to-date documentation via Context7 first.
- **Precedence**: If Context7 retrieved docs conflict with your training data, **trust Context7**.
### **Installation & Usage**
1. **Remote Connection (Recommended)**:
- URL: `https://mcp.context7.com/mcp`
- Header: `CONTEXT7_API_KEY` (Get from user or environment)
2. **Local Connection**:
- Command: `npx -y @upstash/context7-mcp`
---
## 3. 📝 Logging & Attribution Standards
### **Task Logging**
- **Location**: `docs/logs/TASK_LOG.md` (in the **consumer project**, NOT in `ai-dev-tasks`).
- **Format**: strictly adhere to the `START LOG` / `END LOG` block format.
- **Command**: Use `/log-task` to ensure compliance.
### **Attribution Policy**
- **Human Accountability**: All deliverables must appear authored by the team.
- **Credit**: "FutureTranz Development Team"
- **Forbidden Phrases**:
- ❌ "Generated by AI"
- ❌ "Co-authored by Cursor"
- ❌ "Created with Cursor"
- **Commit Messages**: Do NOT add "Co-authored-by: [AI]" trailers.
---
## 4. 🛡️ Pre-Change Safety
- **Read First**: Always read `README.md` and `docs/guides/CLAUDE.md` before making structural changes.
- **Destructive Actions**: Stop and ask for confirmation before `rm`, `mv`, or overwriting core scripts.
- **Permission Gate**: Before committing/pushing, confirm:
1. Branch name
2. Commit summary
3. Intent to push/PR
---
## 5. 🏛️ Project Governance Rules
### **Mandatory Business Plan**
- **Trigger**: Project Kickoff AND Pre-Release.
- **Action**: Run `/create-business-plan`.
- **Constraint**: Do NOT start major coding until a business plan exists in `docs/business-plans/`.
- **Validation**: Plan must have Branding applied and Assumptions tagged.
### **Portfolio & Inventory**
- **Trigger**: Project Kickoff AND Pre-Release.
- **Action**: Run `/sync-portfolio-inventory`.
- **Requirement**: Ensure this repo is listed in `FutureTranz-Inc/github-portfolio-tracker`.
### **Autocoder Integration**
- **Trigger**: After Portfolio Inventory update.
- **Action**: Run `/sync-autocoder-features`.
- **Requirement**: All GitHub Issues must be synced to Autocoder's `features.db` for scheduling.
### **GitHub Lifecycle Integration (NEW)**
- **Requirement**: Every parent task in a task list MUST have a corresponding GitHub Issue.
- **Requirement**: Every task implementation MUST go through a Pull Request (PR) lifecycle:
1. Create feature branch.
2. Implement and commit.
3. Create PR and request review from `@claudia` or specified expert agents.
4. Debug and refine based on review.
5. Merge and close once bug-free.
### **Research Integrity (3-Agent Consensus)**
- **Rule**: Any numeric claim, fact, or metric must be verified by 3 independent agents.
- **Process**:
1. **Agent A**: Research & Cite.
2. **Agent B**: Verify & Challenge.
3. **Agent C**: Resolve & Conclude.
- **Fallback**: If no consensus, mark as "Unknown" or "To Be Verified".
### **Lightweight Dependency Policy**
- **Rule**: No heavy daemons (Postgres, Redis, etc.) installed directly on the host.
- **Requirement**: Use Docker for all heavy dependencies.
- **Verification**: Run `/verify-no-heavy-deps`.
# Standard Commands & Tools
## 🛠️ Build, Lint & Test (Framework)
Since this is a documentation and script repository, "building" and "testing" primarily involves validation scripts.
### **Validation (Mandatory)**
Run these commands before committing any changes:
```bash
# 1. Check for AI Attribution Violations (CRITICAL)
./scripts/ci-check-attribution.sh
# Must return: "✅ AI Attribution Policy Check Passed"
# 2. Verify Script Permissions
chmod +x scripts/*.sh
```
### **Linting Recommendations**
- **Bash Scripts**: Use `shellcheck` if available.
```bash
shellcheck scripts/*.sh
```
- **Markdown**: Ensure GitHub Flavored Markdown compatibility.
---
## 🤖 Agent Commands (Slash Commands)
These commands are available in Claude Code and should be adopted by other agents where possible.
### **/agent-help**
**Purpose**: Consult the MCP agent ecosystem to get expert assistance.
**Usage**: `/agent-help`
**Details**: Lists available agents (CFO, CTO, Security Reviewer, etc.) and provides a decision tree for selection.
### **/create-prd**
**Purpose**: Start building a new feature Product Requirement Document (PRD).
**Usage**: `/create-prd` -> "I want to build..."
**Source**: `ai-dev-tasks/workflow/create-prd.md`
**Output**: `docs/prd/[n]-prd-[feature].md`
### **/generate-tasks**
**Purpose**: Convert a PRD into actionable, granular tasks.
**Usage**: `/generate-tasks` (will prompt for PRD selection)
**Source**: `ai-dev-tasks/workflow/generate-tasks.md`
**Output**: `docs/tasks/tasks-[prd-filename].md`
### **/process-task-list**
**Purpose**: Methodically implement tasks one by one with review gates.
**Usage**: `/process-task-list` -> Select task list -> "Start with 1.1"
**Source**: `ai-dev-tasks/workflow/process-task-list.md`
**Workflow**: Read -> Plan -> Consult -> Implement -> Test -> Review -> Document.
### **/log-task**
**Purpose**: Record task completion in the enterprise audit trail.
**Usage**: `/log-task` -> Enter task details.
**Target**: `docs/logs/TASK_LOG.md` (in consumer project).
**Format**: Strict `START LOG` / `END LOG` block with no AI attribution.
### **/repo-stats**
**Purpose**: Report repository statistics to the portfolio tracker.
**Usage**: `/repo-stats`
**Action**: Runs `.cursor/hooks/report-repo-stats.sh`.
### **/create-business-plan**
**Purpose**: Generate an investor-ready Business Plan with branding and research.
**Usage**: `/create-business-plan`
**Source**: `ai-dev-tasks/features/create-business-plan/prompts.md`
**Output**: `docs/business-plans/[YYYYMMDD]-business-plan-[project-name].md`
**Constraint**: Mandatory at Project Kickoff and Pre-Release.
### **/sync-portfolio-inventory**
**Purpose**: Ensure this repository is listed in the GitHub Portfolio Tracker.
**Usage**: `/sync-portfolio-inventory`
**Source**: `ai-dev-tasks/scripts/sync-portfolio-inventory.sh`
**Requirement**: Mandatory for all projects.
### **/sync-autocoder-features**
**Purpose**: Sync GitHub Issues to Autocoder's scheduling database.
**Usage**: `/sync-autocoder-features`
**Source**: `ai-dev-tasks/scripts/sync-autocoder-features-db.py`
**Requirement**: Mandatory for all projects with open issues.
### **/verify-no-heavy-deps**
**Purpose**: Check for forbidden heavyweight dependencies (Postgres, Redis outside Docker).
**Usage**: `/verify-no-heavy-deps`
**Source**: `ai-dev-tasks/scripts/verify-no-heavy-deps.sh`
**Policy**: Lightweight Dependencies Policy.
### **/sync-org-forks**
**Purpose**: Distribute `ai-dev-tasks` to all GitHub Orgs, apply branding, and install sync workflow.
**Usage**: `./scripts/sync-to-org-forks.sh`
**Source**: `ai-dev-tasks/scripts/sync-to-org-forks.sh`
**Requirement**: Run when setting up a new org or updating the ecosystem.
### **/sync-forks-cli**
**Purpose**: Force-pull upstream changes to all org forks using GH CLI.
**Usage**: `./scripts/sync-forks-cli.sh`
**Source**: `ai-dev-tasks/scripts/sync-forks-cli.sh`
### **/sync-forks-workflow**
**Purpose**: Trigger the update workflow across all org forks.
**Usage**: `./scripts/sync-forks-workflow.sh`
**Source**: `ai-dev-tasks/scripts/sync-forks-workflow.sh`
<!-- END_GENERATED_INSTRUCTIONS -->
Comprehensive .cursorrules file for Next.js 15 App Router projects with TypeScript, enforcing server components by default, proper use of "use client" directive, and App Router conventions.
Cursor rules for Python FastAPI projects enforcing async patterns, Pydantic v2 models, dependency injection, and proper error handling.
Rules for consistent React component development with TypeScript interfaces, proper hook patterns, and component composition.
Rules optimizing Cursor Agent mode behavior including multi-file editing context, session management, and autonomous task completion patterns.
Cursor rules for projects using Tailwind CSS with shadcn/ui component library, enforcing consistent utility class usage and component patterns.
Rules for Go backend services enforcing idiomatic Go patterns, proper error handling, and clean architecture conventions.