bc-code-intelligence-mcp
Version:
BC Code Intelligence MCP Server - Complete Specialist Bundle with AI-driven expert consultation, seamless handoffs, and context-preserving workflows
241 lines (239 loc) ā¢ 11.3 kB
JavaScript
/**
* Core BC Knowledge MCP Tools
*
* The fundamental 8-tool interface for Business Central development assistance.
* These tools provide the primary interface for accessing BC knowledge, specialists,
* and workflows.
*/
/**
* Tool name constants for type safety
*/
export const STREAMLINED_TOOL_NAMES = {
FIND_BC_KNOWLEDGE: 'find_bc_knowledge',
ASK_BC_EXPERT: 'ask_bc_expert',
ANALYZE_AL_CODE: 'analyze_al_code',
GET_BC_TOPIC: 'get_bc_topic',
START_BC_WORKFLOW: 'start_bc_workflow',
ADVANCE_WORKFLOW: 'advance_workflow',
GET_WORKFLOW_HELP: 'get_workflow_help',
GET_BC_HELP: 'get_bc_help'
};
/**
* Core 8-tool interface for BC Code Intelligence MCP Server
*/
export const streamlinedTools = [
{
name: 'find_bc_knowledge',
description: 'Search BC knowledge topics, find specialists, or discover workflows. Use this when users want to find information about BC development.',
inputSchema: {
type: 'object',
properties: {
query: {
type: 'string',
description: 'Search query or question about BC development'
},
search_type: {
type: 'string',
enum: ['topics', 'specialists', 'workflows', 'all'],
description: 'Type of search to perform',
default: 'all'
},
bc_version: {
type: 'string',
description: 'Business Central version (e.g., "BC22", "BC20")'
},
limit: {
type: 'number',
description: 'Maximum number of results to return',
default: 10
}
},
required: ['query']
}
},
{
name: 'ask_bc_expert',
description: `DIRECT SPECIALIST CONSULTATION: Connect with BC specialists who respond immediately as themselves, not as AI assistants describing roleplay. When user says "Talk to Sam" or "Ask Dean about debugging", you get authentic specialist expertise directly. CRITICAL: Do not explain roleplay or offer menus - respond AS the specialist would respond. USE FOR: "Talk to Sam", "Ask Dean about debugging", "I took over this BC app", "Help with performance issues". DO NOT USE FOR WORKFLOWS.
š§ **AL/BC Platform Constraints**: All specialist responses MUST respect Business Central and AL language limitations:
ā¢ Security: AL permission objects, user groups, BC security framework - NOT external auth systems
ā¢ UX: AL page/report constraints - BC controls rendering, NOT custom CSS/HTML
ā¢ Performance: AL optimization, table design, BC server constraints - NOT generic frameworks
ā¢ API: BC API pages, web services, AL integration - NOT generic REST frameworks
ā¢ Always prioritize AL language capabilities and BC platform limitations over generic programming approaches
š¤ **AUTONOMOUS AGENT MODE**: Set autonomous_mode=true for GitHub Coding Agents and autonomous workflows. Returns structured, actionable JSON with: action_plan (executable steps), confidence scores, blocking_issues, and alternatives. Use for Issue ā PR automation.`,
inputSchema: {
type: 'object',
properties: {
question: {
type: 'string',
description: 'Your specific question or challenge about BC development'
},
context: {
type: 'string',
description: 'Optional context about your situation, code, or project'
},
preferred_specialist: {
type: 'string',
description: 'Optional: specific specialist to consult (will auto-detect if not provided)'
},
autonomous_mode: {
type: 'boolean',
description: 'Enable autonomous agent mode: returns structured action plan instead of conversational response. For GitHub Coding Agents and automated workflows.',
default: false
}
},
required: ['question']
}
},
{
name: 'analyze_al_code',
description: 'Analyze AL code for issues, patterns, and improvements. Includes workspace analysis and workflow recommendations. AUTONOMOUS AGENT MODE: Use operation="validate" for automated code validation with compliance checks and auto-fix suggestions. Use operation="suggest_fixes" for code transformation recommendations.',
inputSchema: {
type: 'object',
properties: {
code: {
type: 'string',
description: 'AL code to analyze (or "workspace" to analyze current workspace)'
},
analysis_type: {
type: 'string',
enum: ['performance', 'quality', 'security', 'patterns', 'comprehensive'],
description: 'Type of analysis to perform',
default: 'comprehensive'
},
operation: {
type: 'string',
enum: ['analyze', 'validate', 'suggest_fixes'],
description: 'Analysis operation mode: "analyze" (conversational), "validate" (compliance check + auto-fixes), "suggest_fixes" (code transformations)',
default: 'analyze'
},
bc_version: {
type: 'string',
description: 'Business Central version for version-specific analysis'
},
suggest_workflows: {
type: 'boolean',
description: 'Include workflow recommendations based on analysis',
default: true
}
},
required: ['code']
}
},
{
name: 'get_bc_topic',
description: 'Get detailed content for a specific BC knowledge topic with examples and best practices.',
inputSchema: {
type: 'object',
properties: {
topic_id: {
type: 'string',
description: 'Unique topic identifier (e.g., "sift-technology-fundamentals")'
},
include_samples: {
type: 'boolean',
description: 'Include companion AL code samples if available',
default: true
}
},
required: ['topic_id']
}
},
{
name: 'start_bc_workflow',
description: 'STRUCTURED WORKFLOWS: Start systematic BC development workflows for complex, multi-phase processes. USE FOR STRUCTURED PROCESSES like "optimize my code systematically", "conduct architecture review", "security audit", "performance analysis". DO NOT USE FOR SIMPLE CONVERSATIONS - if user wants to "talk to" or "ask" a specialist, use ask_bc_expert instead. š¤ AUTONOMOUS AGENT MODE: Set execution_mode="autonomous" for GitHub Coding Agents to get next action instead of interactive prompts. Supports checkpoint_id for multi-session execution.',
inputSchema: {
type: 'object',
properties: {
workflow_type: {
type: 'string',
enum: ['new-bc-app', 'enhance-bc-app', 'review-bc-code', 'debug-bc-issues', 'modernize-bc-code', 'onboard-developer', 'upgrade-bc-version', 'add-ecosystem-features', 'document-bc-solution'],
description: 'Type of workflow to start'
},
context: {
type: 'string',
description: 'Project context, code location, or description of what needs to be worked on'
},
bc_version: {
type: 'string',
description: 'Business Central version'
},
execution_mode: {
type: 'string',
enum: ['interactive', 'autonomous'],
description: 'Execution mode: "interactive" (human-in-loop with prompts) or "autonomous" (returns structured next action for automated workflows)',
default: 'interactive'
},
checkpoint_id: {
type: 'string',
description: 'Resume from saved workflow checkpoint (for multi-session execution). Enables stateful workflow progression across multiple invocations.'
},
additional_info: {
type: 'object',
description: 'Additional context specific to the workflow type'
}
},
required: ['workflow_type', 'context']
}
},
{
name: 'advance_workflow',
description: 'Progress to the next phase of an active workflow with your results and feedback.',
inputSchema: {
type: 'object',
properties: {
workflow_id: {
type: 'string',
description: 'ID of the workflow session to advance'
},
phase_results: {
type: 'string',
description: 'Results from the current phase, decisions made, or feedback'
},
next_focus: {
type: 'string',
description: 'Optional: specific focus area for the next phase'
}
},
required: ['workflow_id', 'phase_results']
}
},
{
name: 'get_workflow_help',
description: 'Get guidance for current workflow phase, check status, or understand next steps.',
inputSchema: {
type: 'object',
properties: {
workflow_id: {
type: 'string',
description: 'ID of the workflow session (optional - will list active workflows if not provided)'
},
help_type: {
type: 'string',
enum: ['status', 'guidance', 'next-steps', 'methodology'],
description: 'Type of help to provide',
default: 'guidance'
}
}
}
},
{
name: 'get_bc_help',
description: 'Meta-tool that analyzes your current context and suggests what BC tools or workflows to use next. Use when unsure what to do.',
inputSchema: {
type: 'object',
properties: {
current_situation: {
type: 'string',
description: 'Description of your current situation, challenge, or what you\'re trying to accomplish'
},
workspace_context: {
type: 'string',
description: 'Optional: information about your current AL workspace or project'
}
},
required: ['current_situation']
}
}
];
//# sourceMappingURL=core-tools.js.map