Vibecheck MCP Server — 103 Tools with Risk Policies

// ALL 103 VIBECHECK TOOLS

READ 73 tools

Read agent_gate_history Get gate evaluation history (blocked operations) Read agent_intent_declare Declare an intent before executing operations Read agent_runtime_init Initialize the agent runtime for a project Read agent_tools_list List all registered agent runtime tools Read agent_transcript_get Get the execution transcript for a run Read context_conventions Get project coding conventions Read context_for_file Get relevant context for editing a specific file Read context_gather Gather relevant context for a code generation task Read context_structure Get project structure overview Read docguard_evaluate Evaluate a markdown document against DocGuard rules (duplicate detection, quality checks) Read docguard_find_canonical Find the canonical document for a given topic Read docguard_quick_check Quick check for AI-generated documentation slop Read docguard_scan Scan project and update the documentation registry Read docguard_stats Get documentation registry statistics Read docguard_suggest_location Suggest where new documentation content should be placed Read excessive-claims Warn when change has too many unverified claims Read finding_explanation Get a detailed explanation of why a specific finding type matters and how to fix it. Use this when explaini... Read finding_types_list List all finding types categorized by severity (blockers, warnings, info). Read firewall_evaluate Evaluate code through the agent firewall Read firewall_extract_claims Extract verifiable claims from code Read firewall_get_mode Get the current firewall mode Read firewall_get_tier Get the confidence tier for a firewall rule (hard_block/soft_block/warn) Read firewall_quick_check Quick hallucination check without full evidence resolution Read firewall_should_block Check if a confidence tier should result in a block Read firewall_status Get current firewall status and statistics Read firewall_tier_stats Get firewall statistics broken down by confidence tier Read forge_analyze Run comprehensive project analysis to understand codebase structure, patterns, and conventions. Read forge_contract Get the AI Contract defining allowed and forbidden actions for this project. Read forge_memory_query Search the enhanced memory for relevant information, past conversations, insights, and lessons learned. Read forge_memory_record Record a conversation, decision, or learning in the enhanced memory system. Call this to help Forge learn f... Read forge_memory_stats Get statistics about the enhanced memory system. Read forge_memory_timeline Get the project timeline showing major events and milestones. Read forge_phase Analyze and detect the current project phase using 8 weighted signals. Read forge_rule_feedback Provide feedback on a generated rule to help Forge learn and improve. Read forge_rules List all generated AI context rules with their categories and impact scores. Read forge_status Get current Forge status including phase, rule count, and last update time. Read ghost-env Block usage of undeclared environment variables Read ghost-file Block references to non-existent files Read ghost-route Block references to non-existent API endpoints Read glossary_list List all VibeCheck glossary terms and their brief descriptions. Read glossary_lookup Look up a VibeCheck term definition. Use this when users encounter unfamiliar terminology like Truthpack, S... Read glossary_search Search the VibeCheck glossary for terms matching a query. Read intent_check Check if a specific action is allowed under current intent Read intent_declare Declare the intended scope and operations for upcoming actions Read intent_extend Extend the current intent with additional permissions Read intent_get Get the currently active intent declaration Read intent_history Get history of declared intents Read intent_restrict Restrict the current intent with narrower permissions Read isl_answer_questions Provide answers to open questions from a previous translation. After calling isl_translate, use this t... Read isl_print Convert a valid JSON AST to canonical ISL text. The output is deterministic: same AST always produces ... Read isl_repair_ast Attempt to auto-repair a JSON AST that has validation errors. Returns the repaired AST and ISL, or lis... Read isl_translate Translate natural language to a verified ISL specification. This is the CORRECT way to create ISL specs: 1... Read isl_validate_ast Validate a JSON AST to ensure it Read low-confidence Warn about claims with low confidence Read middleware_intent_history Get intent validation history Read middleware_traces Get recent operation traces for debugging Read onboarding_status Check onboarding progress and get guidance on next steps. Use this to help new users get started with VibeC... Read prompt_get_task Get detailed prompt for a specific planned task Read prompt_quick_check Quick check if a prompt has blocking issues Read prompt_templates Get available prompt templates Read prompt_verify Verify a prompt for potential issues before sending to AI Read quota_info Get quota and feature information for a pricing tier. Use this to explain what features are available. Read quota_status Get current usage quota status for the account Read truthpack_env Get environment variables from truthpack Read truthpack_query Query the truthpack for specific information Read truthpack_routes Get API routes from truthpack Read truthpack_validate Validate truthpack against current codebase Read validation_code Validate code for correctness and style (counts toward scan quota) Read validation_detect_drift Detect drift between truthpack and codebase Read validation_detect_hallucinations Guardrail: detect potential hallucinations in code (phantom routes, mock data, ghost env). Same as extensio... Read validation_verify Verify a claim against multiple sources Read validation_verify_imports Verify all imports in code are valid Read vibecoder_services List VibeCoder services and their CLI commands. Use when guiding users: Health Check (doctor), Guardrail (s...

WRITE 22 tools

Write context_update_embeddings Update context embeddings for semantic search Write docguard_config Get or update DocGuard configuration Write firewall_lockdown Enter lockdown mode - blocks ALL write operations Write firewall_resolve_evidence Resolve evidence for claims Write firewall_set_mode Set the firewall mode (observe/enforce/lockdown) Write firewall_unblock_plan Get a plan to unblock a firewall rejection Write forge_generate Generate or update AI context rules using the Forge engine. Creates optimized rules for Cursor, Windsurf, a... Write ghost-import Block imports that cannot be verified Write ghost-type Warn when referencing undefined types Write hook_file_write Validate file write operation before committing Write hook_post_generation Validate generated code for hallucinations and issues Write isl_generate Generate an ISL behavior contract from a natural language prompt. Use this BEFORE implementing features to ... Write lock_check Check if a file or folder is locked by VibeCheck Write lock_list Get all files and folders currently locked by VibeCheck Write middleware_export_traces Export all traces for analysis Write prompt_plan_task Break down a complex task into smaller, verifiable sub-tasks Write register_convention Register a coding convention Write register_endpoint Register an API endpoint in the truthpack. Use this BEFORE creating a new route to prevent ghost-route errors. Write register_env_var Register an environment variable in the truthpack. Use this BEFORE using a new env var to prevent ghost-env... Write register_pattern Register a new code pattern for the AI to follow Write register_type Register a TypeScript type in the truthpack contracts Write truthpack_generate Generate or refresh the truthpack by scanning the codebase

DESTRUCTIVE 3 tools

Destructive forge_reset Reset Forge state and clear generated files. Use with caution. Destructive intent_clear Clear the current intent declaration (allows all operations) Destructive middleware_clear_traces Clear all stored traces

EXECUTE 5 tools

Execute agent_graph_run Run a standard Plan -> Execute -> Verify -> Summarize graph Execute agent_tool_execute Execute a registered tool with gate checking Execute hook_pre_generation Run pre-generation checks and get enhanced context before code generation Execute isl_from_task Analyze a task and generate both an intent declaration AND an ISL behavior spec. Use this at the START of a... Execute prompt_build Build an enhanced prompt with truthpack context injection

// RUNNING THIS SERVER

The managed route: connect Vibecheck through the PolicyLayer gateway — every tool call above is checked against your policy before it runs, with a full audit log.

DIRECT INSTALL (UNMANAGED) npx -y @vibecheckdev/vibecheck-mcp

// FAQ

How many tools does the Vibecheck MCP server have? +

The Vibecheck MCP server exposes 103 tools across 4 categories: Read, Write, Destructive, Execute.

How do I enforce policies on Vibecheck tools? +

Route the Vibecheck server through the PolicyLayer gateway. Define allow, deny, or approval rules per tool in the dashboard — they are enforced on every call before it reaches the server.

What risk categories do Vibecheck tools fall into? +

Vibecheck tools are categorised as Read (73), Write (22), Destructive (3), Execute (5). Each category has a recommended default policy.

Let agents act without letting them run wild.

Route your MCP servers through PolicyLayer and every tool call is checked against your policy before it runs — allow, deny, or require approval. Per-identity grants. Full audit log. Live in minutes.

SECURE YOUR MCP →

Free to start. No card required.

4,600+ MCP servers and 31,000+ tools scanned and risk-classified.