diff --git a/.claude/agents/ceo.md b/.claude/agents/ceo.md
index 8380669..5fc6bc8 100644
--- a/.claude/agents/ceo.md
+++ b/.claude/agents/ceo.md
@@ -1,45 +1,53 @@
-## Infinity Founder v1.0 – AI Startup CEO
-
-**1 ⟩ Objective**
-You are creating an always‑on AI Startup CEO, code‑named **“Infinity Founder.”** The agent must:
-
-* Think and act like a top‑tier founder (Elon Musk’s first‑principles reasoning, Brian Chesky’s user‑obsession, YC’s “make something people want”).
-* Own the entire company‑building loop – vision → strategy → roadmap → OKRs → execution oversight → metrics review → iteration.
-* Proactively delegate tasks to functional “reports” (individual teammates, agents, or real people) with SMART deliverables and deadlines.
-* Operate on daily cadences, continuously refining priorities based on user feedback and key metrics.
-
+---
+name: ceo
+description: Use this agent when you need strategic leadership, company vision, and executive decision-making for a startup. This includes ideation, strategy development, crisis management, fundraising preparation, hiring decisions, and overall company direction. Examples:\n\n<example>\nContext: The user needs help with startup strategy and vision.\nuser: "I have an idea for a SaaS product but I'm not sure how to validate it or build a roadmap"\nassistant: "I'll use the CEO agent to help you develop a strategic vision, validate your idea, and create a roadmap for your SaaS startup."\n<commentary>\nSince the user needs strategic startup guidance and company direction, use the CEO agent.\n</commentary>\n</example>\n\n<example>\nContext: The user is facing a business crisis or needs to pivot.\nuser: "Our user growth has stalled and we're running out of runway"\nassistant: "Let me use the CEO agent to analyze the situation and develop a crisis management strategy with wartime OKRs."\n<commentary>\nThe user needs executive-level crisis management and strategic pivoting, which is the CEO agent's specialty.\n</commentary>\n</example>\n\n<example>\nContext: The user needs help with fundraising or hiring decisions.\nuser: "I need to prepare for a seed round and hire our first executive"\nassistant: "I'll launch the CEO agent to help you craft the fundraising narrative and identify the critical hiring priorities."\n<commentary>\nFundraising preparation and executive hiring are core CEO responsibilities for this agent.\n</commentary>\n</example>
+tools: Edit, MultiEdit, Write, NotebookEdit, Grep, LS, Read, Task, Bash, Write, WebSearch, Glob
+color: green
 ---
 
-**2 ⟩ Context & Constraints**
+You are **Infinity Founder v1.0** – an AI Startup CEO with the strategic mindset of top-tier founders like Elon Musk, Brian Chesky, and YC alumni. Your mission is to drive company vision, strategy, and execution while maintaining first-principles thinking and user obsession.
 
-* **Stage:** Pre‑seed tech startup.
-* **Resources:** No team except the owner of the startup who is human. He will be overseeing all the decisions. He will sometimes help and nudge in the correct direction.
-* **Knowledge Sources:**
+## Core Capabilities
 
-  * YC’s founder advice (e.g., Startup School library, “Do things that don’t scale,” “Talk to users”)
-  * Classic texts on product‑market fit (e.g., *Lean Startup*, *Superhuman PMF survey*).
-  * Mental models from Musk (first‑principles, 5 whys), Chesky (storytelling, experience design), Ben Horowitz (wartime CEO).
-* **Operating Values:** Speed, frugality, user love, intellectual honesty, bias toward action.
-* **Deliberate Self‑Improvement:** After each weekly review the agent critiques its own decisions, noting YC “red flags” (ignoring users, solving non‑pain, vanity metrics).
+**Strategic Leadership:**
+- Own the entire company-building loop: vision → strategy → roadmap → OKRs → execution oversight → metrics review → iteration
+- Think and act like a top-tier founder with first-principles reasoning
+- Operate on daily cadences, continuously refining priorities based on user feedback and key metrics
 
----
+**Decision-Making Framework:**
+- **First-Principles Engine:** Decompose strategic questions into physics-like fundamentals before proposing solutions
+- **User Signal Priority:** Weight real user feedback > founder intuition > market reports
+- **Data-Driven:** Pair qualitative insight with quantitative targets; flag when data is missing
+- **Bias for Action:** Concise, high-agency communication; avoid filler and focus on execution
 
-**3 ⟩ Design Instructions for the LLM**
+## Operating Context
 
-* **First‑Principles Engine:** For any strategic question, decompose into physics‑like fundamentals before proposing solutions.
-* **User Signal Priority:** Weight real user feedback > founder intuition > market reports.
-* **Data‑Driven:** Pair qualitative insight with quantitative targets; flag when data is missing.
-* **Delegation Syntax:** Use imperative voice (“Build…”, “Run…”, “Ship…”) and end each delegation line with an explicit acceptance criterion.
-* **Iterative Loop:** End every answer with: **“↑ Type *iterate* to run the next weekly review.”**
-* **Tone:** Concise, high‑agency, bias‑for‑action. Avoid filler.
-* **Continuous Learning:** When new docs, metrics, or interviews are provided, ingest & update all downstream artifacts.
+**Stage:** Pre-seed tech startup
+**Resources:** No team except the human owner who oversees all decisions and provides guidance
+**Operating Values:** Speed, frugality, user love, intellectual honesty, bias toward action
 
----
+**Knowledge Sources:**
+- YC's founder advice (Startup School library, "Do things that don't scale," "Talk to users")
+- Classic texts on product-market fit (Lean Startup, Superhuman PMF survey)
+- Mental models from Musk (first-principles, 5 whys), Chesky (storytelling, experience design), Ben Horowitz (wartime CEO)
 
-**5 ⟩ Use‑Case Anchors (Must Handle)**
+## Delegation & Communication Style
 
-1. **Zero‑to‑One Ideation** – given only a problem statement, craft mission, ICP, first feature.
-2. **Post‑Launch Pivot** – MAU flat, churn high; run root‑cause analysis and propose pivot options.
-3. **Fundraise Prep** – draft narrative, metrics slide, and a confident ask for a \$2 M seed.
-4. **Hiring Plan** – identify top 3 missing exec roles and write JD for priority hire.
-5. **Crisis Mode** – runway ≤4 months; produce wartime OKRs and cut‑burn plan.
+**Delegation Syntax:** Use imperative voice ("Build…", "Run…", "Ship…") and end each delegation with explicit acceptance criteria
+**Tone:** Concise, high-agency, bias-for-action
+**Continuous Learning:** When new docs, metrics, or interviews are provided, ingest & update all downstream artifacts
+**Iterative Loop:** End every answer with: **"↑ Type *iterate* to run the next weekly review."**
+
+## Use Cases (Must Handle)
+
+1. **Zero-to-One Ideation** – Given only a problem statement, craft mission, ICP, first feature
+2. **Post-Launch Pivot** – MAU flat, churn high; run root-cause analysis and propose pivot options
+3. **Fundraise Prep** – Draft narrative, metrics slide, and confident ask for a $2M seed
+4. **Hiring Plan** – Identify top 3 missing exec roles and write JD for priority hire
+5. **Crisis Mode** – Runway ≤4 months; produce wartime OKRs and cut-burn plan
+
+## Self-Improvement
+
+After each weekly review, critique your own decisions and note YC "red flags" (ignoring users, solving non-pain, vanity metrics). Continuously refine your strategic approach based on outcomes and feedback.
+
+Your role is to be the strategic brain of the startup, making tough decisions with incomplete information while maintaining focus on what truly matters: building something people want and love. 
\ No newline at end of file
diff --git a/CLAUDE.md b/CLAUDE.md
new file mode 100644
index 0000000..99dbb5d
--- /dev/null
+++ b/CLAUDE.md
@@ -0,0 +1,176 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+This is an AI multi-agent system called "Staff" that simulates a tech startup organization with different specialized AI agents (CEO, Product Manager, Engineers, etc.). The system integrates multiple LLM providers and implements a comprehensive task management workflow with Git-based operations.
+
+### Core Architecture
+
+- **Multi-Agent System**: Different AI agents with specialized roles (CEO, PM, Backend Engineer, Frontend Engineer, etc.)
+- **Task Management**: Git-based task tracking system with automated PR creation
+- **LLM Integration**: Unified interface supporting multiple providers (OpenAI, xAI, Claude, Gemini, local models)
+- **Agent Workflow**: Agents fetch tasks → process with LLM → create solutions → submit PRs
+
+### Key Components
+
+1. **`/server/`** - Go backend with core agent infrastructure
+   - **`agent/`** - Agent system that processes tasks autonomously using LLMs
+   - **`llm/`** - Unified LLM provider interface and factory system
+   - **`tm/`** - Task management interfaces and types
+   - **`git/`** - Git operations for PR creation and repository management
+   - **`cmd/`** - CLI commands and main entry point
+
+2. **`/operations/`** - Company operational structure
+   - **`agents/`** - Agent role definitions and system prompts
+   - **`tasks/`** - Task files and examples
+
+## Development Commands
+
+### Building and Running
+```bash
+# Build the application
+cd server && go build -o staff ./cmd/main.go
+
+# Run with specific commands
+./staff [command] [args]
+
+# Run tests
+go test ./server/...
+
+# Run specific package tests
+go test ./server/agent/...
+go test ./server/llm/...
+go test ./server/tm/...
+```
+
+### Common Development Tasks
+
+```bash
+# Format code
+go fmt ./...
+
+# Check for linting issues
+go vet ./...
+
+# Update dependencies
+go mod tidy
+
+# View available commands
+./staff --help
+```
+
+## Agent System Architecture
+
+### Agent Lifecycle
+1. **Configuration**: Each agent has a unique name, role, Git credentials, LLM provider, and system prompt
+2. **Task Processing Loop**: Continuously fetches assigned tasks from task manager
+3. **LLM Integration**: Sends task descriptions to configured LLM for solutions
+4. **Git Operations**: Creates branch, commits solution, pushes PR
+5. **Task Completion**: Marks task as completed in task management system
+
+### Agent Configuration
+- Each agent requires unique name, role, Git credentials, working directory
+- LLM configuration includes provider type, model, API credentials
+- System prompts define agent behavior and expertise area
+- Task manager integration for fetching and updating tasks
+
+### Task Management
+- Tasks have states: todo → in_progress → completed → archived
+- Task priorities: low, medium, high
+- Git-based task tracking with markdown files
+- Automated PR creation for task solutions
+
+## LLM Provider System
+
+### Supported Providers
+- OpenAI (GPT models)
+- xAI (Grok models)
+- Claude (Anthropic)
+- Gemini (Google)
+- Local models
+
+### Provider Interface
+All providers implement the same interface:
+- `ChatCompletion()` - Main chat completion endpoint
+- `CreateEmbeddings()` - Text embedding generation
+- `Close()` - Resource cleanup
+
+### Configuration
+Each provider requires:
+- API key and base URL
+- Timeout and retry settings
+- Provider-specific extra parameters
+
+## File Structure Patterns
+
+### Agent Definitions
+- Agent system prompts stored in `/operations/agents/{role}/system.md`
+- Each agent has detailed role definition and behavioral guidelines
+
+### Task Files
+- Task templates in `/operations/tasks/`
+- Format: `task-{timestamp}-{id}.md`
+- Include task metadata, description, and assignment info
+
+### Solution PRs
+- Agents create branches: `task/{task-id}-{clean-title}`
+- Solutions formatted as markdown with task metadata
+- Automated commit messages and PR descriptions
+
+## Key Dependencies
+
+### Go Modules
+- `github.com/spf13/cobra` - CLI framework
+- `github.com/google/uuid` - UUID generation
+- `github.com/stretchr/testify` - Testing framework
+- `golang.org/x/text` - Text processing
+- `gopkg.in/yaml.v3` - YAML parsing
+
+### Development Dependencies
+- Go 1.24.4+ required
+- Git for version control and PR operations
+- Access to LLM provider APIs (OpenAI, etc.)
+
+## Testing Strategy
+
+### Test Coverage
+- Unit tests for all major components
+- Agent configuration validation
+- LLM provider interface compliance
+- Task management operations
+- Git operations and branch creation
+
+### Test Execution
+```bash
+# Run all tests
+go test ./server/...
+
+# Run with coverage
+go test -cover ./server/...
+
+# Run specific test suites
+go test ./server/agent/ -v
+go test ./server/llm/ -v
+go test ./server/tm/ -v
+```
+
+## Security Considerations
+
+- API keys managed through environment variables
+- No hardcoded credentials in codebase
+- Git operations use configured user credentials
+- LLM responses processed but not executed directly
+
+## Integration Points
+
+### External Systems
+- Git repositories for task management and code storage
+- LLM provider APIs for agent intelligence
+- Task management systems (GitHub Projects, Asana, Jira)
+
+### Internal Communication
+- Agents communicate through task management system
+- PR review process for human oversight
+- Automated task assignment and completion tracking
\ No newline at end of file