@pcircle/memesh 2.10.1 → 3.0.0

package/dist/mcp/resources/best-practices.md

@@ -1,53 +0,0 @@
-# MeMesh Best Practices
-
-**Version: 2.8.0
-
-## ✅ Write Precise Requests
-
-- Specify files, modules, and constraints
-- State desired behavior and edge cases
-- Mention the tech stack when relevant
-
-**Good**:
-```
-"Refactor src/auth/session.ts to remove duplication and add tests"
-```
-
-**Avoid**:
-```
-"Fix auth"
-```
-
-## ✅ Use Capability Routing (Default)
-
-Let `buddy-do` select the best capability. You do not need to target internal roles.
-
-```
-buddy-do "review src/api/auth.ts for security issues"
-```
-
-## ✅ Break Down Large Features
-
-Use `memesh-generate-tests` before large changes:
-```
-memesh-generate-tests "add email-based login with rate limiting"
-```
-
-Then execute tasks one by one.
-
-## ✅ Leverage Project Memory
-
-Use memory before large refactors or architectural changes:
-```
-buddy-remember "why we chose SQLite"
-```
-
-## ✅ Prefer Smaller, Testable Chunks
-
-- Aim for 2-5 minute tasks
-- Validate with tests before moving on
-- Use clear commit boundaries in your VCS
-
----
-
-**Remember**: Specific, focused requests yield the best outcomes.

package/dist/mcp/resources/buddy-do-skill.md

@@ -1,84 +0,0 @@
-# buddy-do Tool Reference
-
-## Overview
-
-`buddy-do` is the primary task execution tool in MeMesh. It analyzes task descriptions, applies relevant context from your project memory, and automatically captures task metadata to the Knowledge Graph.
-
-## Basic Usage
-
-```typescript
-await callTool('buddy-do', {
-  task: 'Implement user authentication because we need to secure the API so that only authorized users can access resources'
-});
-```
-
-## Automatic Memory Tracking (Phase 0.6)
-
-The buddy-do tool now automatically captures task metadata to the Knowledge Graph:
-
-### What Gets Recorded
-
-When you call `buddy-do`, the following metadata is automatically extracted and recorded:
-
-1. **Task Goal**: The main objective (extracted from first sentence or "to X" pattern)
-2. **Reason**: Why this task is needed (extracted from "because X" or "so that X" patterns)
-3. **Expected Outcome**: What should happen (extracted from "should X" or "will X" patterns)
-
-### Example
-
-```typescript
-await callTool('buddy-do', {
-  task: 'Implement user authentication because we need to secure the API so that only authorized users can access resources'
-});
-
-// Automatically records to Knowledge Graph:
-// - Goal: "Implement user authentication"
-// - Reason: "we need to secure the API"
-// - Expected Outcome: "only authorized users can access resources"
-```
-
-### Query Past Tasks
-
-Use `buddy-remember` to recall past tasks and their context:
-
-```typescript
-await callTool('buddy-remember', {
-  query: 'authentication implementation decisions'
-});
-```
-
-## Task Description Best Practices
-
-To maximize the effectiveness of automatic memory tracking, structure task descriptions to include:
-
-1. **Clear goal** (what needs to be done)
-2. **Reason** (why it's needed) - use "because" or "so that"
-3. **Expected outcome** (what should result) - use "should" or "will"
-
-### Good Examples
-
-```
-"Refactor UserService to use TypeScript because type safety will prevent runtime errors"
-"Add caching layer to API so that response times will improve by 50%"
-"Implement retry logic because network requests should handle transient failures"
-```
-
-### Less Effective Examples
-
-```
-"Fix the bug" - Too vague, no context
-"Update code" - No goal, reason, or outcome specified
-"Implement feature X" - Missing reason and expected outcome
-```
-
-## Integration with Other Tools
-
-`buddy-do` works seamlessly with:
-- **buddy-remember**: Query knowledge captured from previous tasks
-- **memesh-generate-tests**: Create implementation plans that reference past tasks
-
-## See Also
-
-- [Auto-Memory System Guide](../../../docs/guides/auto-memory-system.md)
-- [Complete Usage Guide](./usage-guide.md)
-- [Best Practices](./best-practices.md)

package/dist/mcp/resources/ccb-complete-guide.md

@@ -1,247 +0,0 @@
-# MeMesh - Complete Tool Guide
-
-**For LLMs**: This guide helps you actively use MeMesh's tools. Use tools proactively based on triggers, not just when explicitly requested.
-
-## Quick Reference: When to Use Each Tool
-
-| User Says | Use Tool | Why |
-|-----------|----------|-----|
-| "remember when we..." | `buddy-remember` | Search project memory |
-| "what did we decide..." | `buddy-remember` | Recall past decisions |
-| "why did we choose..." | `buddy-remember` | Find architecture choices |
-| "do this task..." | `buddy-do` | Execute with memory context |
-| "implement X..." | `buddy-do` | Task execution |
-| "generate tests..." | `memesh-generate-tests` | Create test cases |
-
----
-
-## Core Tools (Daily Use)
-
-### buddy-remember: Project Memory Search
-**Triggers**: remember, recall, past, decision, why, when, bug, pattern, architecture
-
-**When to use**:
-- User asks "why did we..."
-- User asks "remember when..."
-- Starting work on existing feature (check past decisions)
-- Before making architectural changes (review previous choices)
-- Debugging (check if similar bug was fixed before)
-
-**Examples**:
-```typescript
-// User: "Why did we use JWT instead of sessions?"
-buddy-remember({ query: "JWT authentication decision" })
-
-// User: "Has this bug happened before?"
-buddy-remember({ query: "login timeout bug" })
-
-// Proactive: Starting auth work
-buddy-remember({ query: "authentication patterns" })
-```
-
-**Best practices**:
-- Search BEFORE implementing (check existing patterns)
-- Use specific keywords from domain (not generic terms)
-- Query when user mentions "we did this before"
-
----
-
-### buddy-do: Smart Task Execution
-**Triggers**: implement, create, build, fix, refactor, add, update
-
-**When to use**:
-- User requests implementation work
-- Task requires routing to appropriate capability
-- Want automatic task metadata capture
-
-**Examples**:
-```typescript
-// User: "Add rate limiting to the API"
-buddy-do({
-  task: "Add rate limiting to API endpoints because we're getting too many requests so that server stays responsive under load"
-})
-
-// Include: what (rate limiting) + why (too many requests) + outcome (stay responsive)
-```
-
-**Best practices**:
-- Include goal + reason + expected outcome
-- Let buddy-do handle routing (don't manually route)
-- Task metadata auto-saved to knowledge graph
-
----
-
-### memesh-create-entities: Store Knowledge
-**Triggers**: Completed major work, made important decision, learned lesson, fixed bug
-
-**When to use** (PROACTIVE):
-- After implementing new feature (store decision rationale)
-- After fixing bug (record root cause)
-- After making architectural choice (document why)
-- User explicitly corrects you (record mistake pattern)
-
-**Examples**:
-```typescript
-// After implementing OAuth
-memesh-create-entities({
-  entities: [{
-    name: "OAuth2 Implementation Decision",
-    entityType: "decision",
-    observations: [
-      "Chose OAuth2 over JWT for better revocation control",
-      "Used authorization code flow for web clients",
-      "Implemented PKCE for mobile apps"
-    ],
-    tags: ["authentication", "oauth", "security"]
-  }]
-})
-
-// After fixing bug
-memesh-create-entities({
-  entities: [{
-    name: "Login Timeout Bug Fix",
-    entityType: "bug_fix",
-    observations: [
-      "Root cause: Redis connection pooling exhaustion",
-      "Solution: Increased pool size to 20 connections",
-      "Prevention: Added connection pool monitoring"
-    ],
-    tags: ["bug", "redis", "timeout"]
-  }]
-})
-```
-
-**Best practices**:
-- Store AFTER completing significant work (not during)
-- Include specific details (not vague summaries)
-- Add relevant tags for future searchability
-
----
-
----
-
-## Learning Tools (Improvement)
-
-### memesh-record-mistake: Error Learning
-**Triggers**: User corrects you, says "you should have...", "why didn't you..."
-
-**When to use** (IMMEDIATE):
-- User explicitly corrects your behavior
-- You violated documented procedure
-- You made assumption instead of asking
-- You skipped validation step
-- User frustrated with repeated mistake
-
-**Examples**:
-```typescript
-// User: "Why did you edit the file without reading it first?"
-memesh-record-mistake({
-  action: "Edited config.ts without reading first",
-  errorType: "procedure-violation",
-  userCorrection: "Must read file before editing",
-  correctMethod: "Use Read tool first, then Edit with exact content",
-  impact: "Broke file formatting, wasted time",
-  preventionMethod: "ALWAYS Read before Edit - no exceptions"
-})
-```
-
-**Best practices**:
-- Record IMMEDIATELY when corrected (don't wait)
-- Be specific about what you did wrong
-- Include concrete prevention method
-
----
-
-## Test Generation
-
-### memesh-generate-tests: AI Test Creation
-**Triggers**: "write tests", "need tests", "generate tests", finished feature without tests
-
-**When to use**:
-- Completed feature implementation (no tests yet)
-- User explicitly requests test generation
-- Following TDD (specification → tests → implementation)
-
-**Examples**:
-```typescript
-// After implementing function
-memesh-generate-tests({
-  code: `
-    function validateEmail(email: string): boolean {
-      const regex = /^[^\\s@]+@[^\\s@]+\\.[^\\s@]+$/;
-      return regex.test(email);
-    }
-  `
-})
-
-// From specification
-memesh-generate-tests({
-  specification: "Email validator should: accept valid emails, reject invalid formats, handle edge cases (empty, null, special chars)"
-})
-```
-
----
-
-## Common Workflows
-
-### Workflow 1: Starting New Feature
-```
-1. buddy-remember({ query: "similar features patterns" })
-   → Check if similar work exists
-2. buddy-do({ task: "Implement X because Y so that Z" })
-   → Execute with metadata capture
-3. memesh-generate-tests({ code: "..." })
-   → Create tests
-4. memesh-create-entities({ ... })
-   → Store implementation decisions
-```
-
-### Workflow 2: Bug Investigation
-```
-1. buddy-remember({ query: "bug error message" })
-   → Check if bug seen before
-2. buddy-do({ task: "Debug X" })
-   → Investigate and fix
-3. memesh-create-entities({ entityType: "bug_fix", ... })
-   → Record root cause and solution
-```
-
----
-
-## Proactive Tool Usage Rules
-
-**ALWAYS use without asking**:
-- `buddy-remember` - Before implementing similar features
-- `memesh-create-entities` - After completing significant work
-- `memesh-record-mistake` - When user corrects you
-
-**SOMETIMES use proactively**:
-- `memesh-generate-tests` - After implementing untested code (offer)
----
-
-## Anti-Patterns (What NOT to Do)
-
-❌ **Don't**:
-- Use `buddy-remember` with generic queries ("authentication")
-- Call `buddy-do` for trivial tasks (use direct tools)
-- Skip `memesh-create-entities` after major work
-- Ignore `memesh-record-mistake` when corrected
-
-✅ **Do**:
-- Use specific queries ("OAuth2 scope design decision")
-- Let `buddy-do` handle complex multi-step tasks
-- Proactively store knowledge after completion
-- Immediately record mistakes for learning
-
----
-
-## Tool Discovery Tips
-
-**For LLMs reading this guide**:
-1. Match user keywords to tool triggers (see Quick Reference)
-2. Use tools proactively (don't wait for explicit requests)
-3. Combine tools in workflows (see Common Workflows)
-4. Store knowledge after work (future you will thank you)
-5. Learn from mistakes (use memesh-record-mistake)
-
-**Remember**: MeMesh tools are meant to be ACTIVELY used, not passively available. When you see trigger keywords or situations, use the appropriate tool to provide better service to users.

package/dist/mcp/resources/examples.md

@@ -1,47 +0,0 @@
-# MeMesh Examples
-
-**Version: 2.8.0
-
-## Example 1: Code Review
-
-**Request**:
-```
-buddy-do "review src/auth/login.ts for security issues and edge cases"
-```
-
-**Result**: A focused review with security checks, edge cases, and actionable fixes.
-
----
-
-## Example 2: Feature Planning
-
-**Request**:
-```
-memesh-generate-tests "add email-based login with rate limiting"
-```
-
-**Result**: A structured plan with TDD-friendly steps and estimated durations.
-
----
-
-## Example 3: Debugging a Failure
-
-**Request**:
-```
-buddy-do "debug why login requests return 500s in production"
-```
-
-**Result**: Root-cause analysis with likely failure paths and verification steps.
-
----
-
-## Example 4: Memory Recall
-
-**Request**:
-```
-buddy-remember "why we chose SQLite for local storage"
-```
-
-**Result**: A concise summary of the decision and its tradeoffs.
-
-

package/dist/mcp/resources/handlers/AgentStatusHandler.d.ts

@@ -1,7 +0,0 @@
-import { ResourceContent } from '../ResourceRegistry.js';
-import { URITemplateParams } from '../URITemplateHandler.js';
-export declare class AgentStatusHandler {
-    handle(params: URITemplateParams): Promise<ResourceContent>;
-    private getAgentCapabilities;
-}
-//# sourceMappingURL=AgentStatusHandler.d.ts.map

package/dist/mcp/resources/handlers/AgentStatusHandler.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"AgentStatusHandler.d.ts","sourceRoot":"","sources":["../../../../src/mcp/resources/handlers/AgentStatusHandler.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AACzD,OAAO,EAAE,iBAAiB,EAAE,MAAM,0BAA0B,CAAC;AAyB7D,qBAAa,kBAAkB;IAOvB,MAAM,CAAC,MAAM,EAAE,iBAAiB,GAAG,OAAO,CAAC,eAAe,CAAC;IAqCjE,OAAO,CAAC,oBAAoB;CAW7B"}

package/dist/mcp/resources/handlers/AgentStatusHandler.js

@@ -1,41 +0,0 @@
-import { NotFoundError } from '../../../errors/index.js';
-const VALID_AGENT_TYPES = [
-    'code-reviewer',
-    'test-writer',
-    'development-butler',
-    'e2e-healing',
-    'knowledge-graph',
-];
-export class AgentStatusHandler {
-    async handle(params) {
-        const agentType = params.agentType;
-        if (!agentType) {
-            throw new Error('Missing required parameter: agentType');
-        }
-        if (!VALID_AGENT_TYPES.includes(agentType)) {
-            throw new NotFoundError(`Unknown agent type: ${agentType}`, 'agent', agentType, { validTypes: VALID_AGENT_TYPES });
-        }
-        const status = {
-            agentType,
-            status: 'active',
-            capabilities: this.getAgentCapabilities(agentType),
-            lastActive: new Date().toISOString(),
-        };
-        return {
-            uri: `ccb://agent/${agentType}/status`,
-            mimeType: 'application/json',
-            text: JSON.stringify(status, null, 2),
-        };
-    }
-    getAgentCapabilities(agentType) {
-        const capabilities = {
-            'code-reviewer': ['review', 'analyze', 'suggest'],
-            'test-writer': ['generate-tests', 'test-analysis'],
-            'development-butler': ['project-setup', 'dependency-management'],
-            'e2e-healing': ['test-healing', 'failure-analysis'],
-            'knowledge-graph': ['memory', 'learning', 'recall'],
-        };
-        return capabilities[agentType] || [];
-    }
-}
-//# sourceMappingURL=AgentStatusHandler.js.map

package/dist/mcp/resources/handlers/AgentStatusHandler.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"AgentStatusHandler.js","sourceRoot":"","sources":["../../../../src/mcp/resources/handlers/AgentStatusHandler.ts"],"names":[],"mappings":"AAGA,OAAO,EAAE,aAAa,EAAE,MAAM,0BAA0B,CAAC;AASzD,MAAM,iBAAiB,GAAG;IACxB,eAAe;IACf,aAAa;IACb,oBAAoB;IACpB,aAAa;IACb,iBAAiB;CACT,CAAC;AASX,MAAM,OAAO,kBAAkB;IAO7B,KAAK,CAAC,MAAM,CAAC,MAAyB;QACpC,MAAM,SAAS,GAAG,MAAM,CAAC,SAAS,CAAC;QAEnC,IAAI,CAAC,SAAS,EAAE,CAAC;YACf,MAAM,IAAI,KAAK,CAAC,uCAAuC,CAAC,CAAC;QAC3D,CAAC;QAED,IAAI,CAAC,iBAAiB,CAAC,QAAQ,CAAC,SAAgB,CAAC,EAAE,CAAC;YAClD,MAAM,IAAI,aAAa,CACrB,uBAAuB,SAAS,EAAE,EAClC,OAAO,EACP,SAAS,EACT,EAAE,UAAU,EAAE,iBAAiB,EAAE,CAClC,CAAC;QACJ,CAAC;QAGD,MAAM,MAAM,GAAgB;YAC1B,SAAS;YACT,MAAM,EAAE,QAAQ;YAChB,YAAY,EAAE,IAAI,CAAC,oBAAoB,CAAC,SAAS,CAAC;YAClD,UAAU,EAAE,IAAI,IAAI,EAAE,CAAC,WAAW,EAAE;SACrC,CAAC;QAEF,OAAO;YACL,GAAG,EAAE,eAAe,SAAS,SAAS;YACtC,QAAQ,EAAE,kBAAkB;YAC5B,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC;SACtC,CAAC;IACJ,CAAC;IAQO,oBAAoB,CAAC,SAAiB;QAC5C,MAAM,YAAY,GAA6B;YAC7C,eAAe,EAAE,CAAC,QAAQ,EAAE,SAAS,EAAE,SAAS,CAAC;YACjD,aAAa,EAAE,CAAC,gBAAgB,EAAE,eAAe,CAAC;YAClD,oBAAoB,EAAE,CAAC,eAAe,EAAE,uBAAuB,CAAC;YAChE,aAAa,EAAE,CAAC,cAAc,EAAE,kBAAkB,CAAC;YACnD,iBAAiB,EAAE,CAAC,QAAQ,EAAE,UAAU,EAAE,QAAQ,CAAC;SACpD,CAAC;QAEF,OAAO,YAAY,CAAC,SAAS,CAAC,IAAI,EAAE,CAAC;IACvC,CAAC;CACF"}

package/dist/mcp/resources/handlers/TaskLogsHandler.d.ts

@@ -1,7 +0,0 @@
-import { ResourceContent } from '../ResourceRegistry.js';
-import { URITemplateParams } from '../URITemplateHandler.js';
-export declare class TaskLogsHandler {
-    handle(params: URITemplateParams): Promise<ResourceContent>;
-    private fetchTaskLogs;
-}
-//# sourceMappingURL=TaskLogsHandler.d.ts.map

package/dist/mcp/resources/handlers/TaskLogsHandler.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"TaskLogsHandler.d.ts","sourceRoot":"","sources":["../../../../src/mcp/resources/handlers/TaskLogsHandler.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AACzD,OAAO,EAAE,iBAAiB,EAAE,MAAM,0BAA0B,CAAC;AAS7D,qBAAa,eAAe;IAOpB,MAAM,CAAC,MAAM,EAAE,iBAAiB,GAAG,OAAO,CAAC,eAAe,CAAC;YA+BnD,aAAa;CAoB5B"}

package/dist/mcp/resources/handlers/TaskLogsHandler.js

@@ -1,35 +0,0 @@
-export class TaskLogsHandler {
-    async handle(params) {
-        const taskId = params.taskId;
-        if (!taskId) {
-            throw new Error('Missing required parameter: taskId');
-        }
-        const logs = await this.fetchTaskLogs(taskId);
-        return {
-            uri: `ccb://task/${taskId}/logs`,
-            mimeType: 'text/plain',
-            text: logs,
-        };
-    }
-    async fetchTaskLogs(taskId) {
-        return `Task Logs - Feature Not Yet Available
-=====================================
-
-Task ID: ${taskId}
-
-Detailed task execution logs are not yet available in this version.
-
-This feature requires:
-- Log capture in task storage
-- Log storage in Cloud storage
-- Integration with task lifecycle tracking
-
-Current workaround:
-- Monitor console logs during task execution
-- Use UIEventBus for progress events
-
-Status: Planned for MeMesh Cloud integration
-`;
-    }
-}
-//# sourceMappingURL=TaskLogsHandler.js.map

package/dist/mcp/resources/handlers/TaskLogsHandler.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"TaskLogsHandler.js","sourceRoot":"","sources":["../../../../src/mcp/resources/handlers/TaskLogsHandler.ts"],"names":[],"mappings":"AAWA,MAAM,OAAO,eAAe;IAO1B,KAAK,CAAC,MAAM,CAAC,MAAyB;QACpC,MAAM,MAAM,GAAG,MAAM,CAAC,MAAM,CAAC;QAE7B,IAAI,CAAC,MAAM,EAAE,CAAC;YACZ,MAAM,IAAI,KAAK,CAAC,oCAAoC,CAAC,CAAC;QACxD,CAAC;QAGD,MAAM,IAAI,GAAG,MAAM,IAAI,CAAC,aAAa,CAAC,MAAM,CAAC,CAAC;QAE9C,OAAO;YACL,GAAG,EAAE,cAAc,MAAM,OAAO;YAChC,QAAQ,EAAE,YAAY;YACtB,IAAI,EAAE,IAAI;SACX,CAAC;IACJ,CAAC;IAgBO,KAAK,CAAC,aAAa,CAAC,MAAc;QACxC,OAAO;;;WAGA,MAAM;;;;;;;;;;;;;;CAchB,CAAC;IACA,CAAC;CACF"}

package/dist/mcp/resources/quick-reference.md

@@ -1,71 +0,0 @@
-# MeMesh Quick Reference
-
-**Version: 2.9.0
-
-## Quick Commands
-
-```bash
-buddy-do "<task description>"         # Execute with memory context (recommended)
-buddy-remember "<query>"             # Recall project memory
-buddy-help                             # Command help
-memesh-generate-tests '{"featureDescription":"..."}'
-```
-
-Note: `hook-tool-use` is internal and used by Claude Code hooks.
-
-## Capability Keywords
-
-| Capability | Keywords | Use For |
-|---|---|---|
-| Code Review | review, audit, quality | Security checks, best practices |
-| Testing | test, coverage, TDD | Unit, integration, E2E testing |
-| Debugging | debug, bug, error | Root-cause analysis |
-| Refactoring | refactor, cleanup | Structural improvements |
-| API/Backend | api, endpoint, backend | API design and backend logic |
-| Frontend/UI | ui, component, frontend | UI, accessibility, UX |
-| Architecture | architecture, design, scalability | System design, tradeoffs |
-| Data/Database | database, query, schema | Query tuning, schema design |
-| Documentation | docs, guide, readme | Technical writing |
-
-## Common Workflows
-
-### Feature Development
-```
-1. memesh-generate-tests "<feature>"
-2. buddy-do "implement <feature>"
-3. buddy-do "write tests for <feature>"
-4. buddy-do "review <feature> for quality"
-```
-
-### Bug Fixing
-```
-1. buddy-do "debug <issue> and find root cause"
-2. buddy-do "fix <issue> and add regression test"
-```
-
-### Performance Review
-```
-1. buddy-do "profile performance for <area>"
-2. buddy-do "optimize slow path in <area>"
-```
-
-## Quick Tips
-
-**DO**:
-- Use clear, specific task descriptions
-- Mention files, constraints, or target behaviors
-- Use `buddy-remember` before large changes
-
-**DON'T**:
-- Use vague descriptions ("fix bug")
-- Skip context when it matters
-
-## Full Documentation
-
-- **Complete Guide**: `@claude-code-buddy://usage-guide`
-- **Examples**: `@claude-code-buddy://examples`
-- **Best Practices**: `@claude-code-buddy://best-practices`
-
----
-
-**Tip**: Use `buddy-remember` to recall past decisions and patterns before starting new work.

package/dist/mcp/resources/usage-guide.md

@@ -1,74 +0,0 @@
-# MeMesh Complete Usage Guide
-
-**Version: 2.8.0
-**Last Updated**: 2026-02-04
-
-## 📖 Overview
-
-MeMesh is a capability-routing system for Claude Code. It focuses on:
-- Smart task routing
-- Project memory
-- Workflow guidance
-- Structured planning
-
-You describe the outcome, and MeMesh selects the best internal capability automatically.
-
-## 🚀 Quick Start
-
-### 1) Run a task
-```
-buddy-do "implement user authentication with JWT"
-```
-
-### 2) Recall project memory
-```
-buddy-remember "authentication decisions"
-```
-
-### 3) Generate a plan
-```
-memesh-generate-tests "add email-based login"
-```
-
-## 🧰 Tool Overview
-
-- **buddy-do**: Execute tasks with memory context
-- **buddy-remember**: Recall project memory and decisions
-- **buddy-help**: Command help and usage
-- **memesh-generate-tests**: Implementation plan generation
-
-Note: `hook-tool-use` is internal and used by Claude Code hooks.
-
-## 🧭 Capability Examples
-
-Use natural language and include intent, constraints, and file names when possible.
-
-```
-"Review src/auth.ts for security and correctness"
-"Debug why login is returning 500s after deploy"
-"Refactor UserService to reduce duplication"
-"Design a REST API for session management"
-"Optimize the database query in reports.sql"
-"Write vitest tests for src/utils/date.ts"
-"Draft README updates for new onboarding flow"
-```
-
-## 🧠 Project Memory
-
-Project memory is updated automatically at key milestones and commit events. Use it to recall prior decisions and patterns:
-```
-buddy-remember "why we chose SQLite"
-buddy-remember "error handling strategy"
-```
-
-## ✅ Best Practices
-
-- Be specific: mention files, constraints, and expected behavior
-- Use `buddy-remember` before large changes
-- Ask for a plan before large features
-
-## 📚 Additional Resources
-
-- **Quick Reference**: `@claude-code-buddy://quick-reference`
-- **Examples**: `@claude-code-buddy://examples`
-- **Best Practices**: `@claude-code-buddy://best-practices`