# asepsafrudin/multiple-mcp-server-by-kimi

Source: https://glimind.com/server/github%2Fasepsafrudin%2Fmultiple-mcp-server-by-kimi
Registry id: `github/asepsafrudin/multiple-mcp-server-by-kimi`
Status: unknown

## Tools (56)

- `audit_log`: Append a structured audit log entry.
- `check_url_status`: Check whether a URL is reachable via HEAD request.
- `decision_triage`: Menganalisis instruksi/tugas dan mengembalikan klasifikasi, tipe routing agen, dan confidence score. Sangat cepat (~1.1s).
- `delete_file`: Delete a file. This operation cannot be undone.
- `document_extract_text`: Extract text from a PDF document sequentially. Can be used by Harvester (Baidu OCR logic placeholder) or agent to read text.
- `document_view_page`: Render a specific page of a PDF document into a high-resolution base64 PNG image. Used for On-Demand Visual Rendering (bypassing OCR resolution degradation).
- `fetch_url`: Fetch content from a URL; optionally strip HTML tags.
- `gemini_generate`: Generate text with Google Gemini.
- `get_env_variable`: Read a non-sensitive environment variable.
- `get_system_info`: Return host system statistics (CPU, memory, OS, Python version).
- `gmail_get_message`: Fetch a Gmail message by ID.
- `gmail_list_messages`: List recent Gmail messages matching an optional query.
- `gmail_send_message`: Send a plain-text email via Gmail.
- `hash_string`: Hash a string using sha256, sha512, or md5 (md5 for checksums only).
- `hindsight_reflect_and_learn`: Derive learning from a past experience and save it as advice for similar future tasks.
- `hindsight_search_advice`: Retrieve relevant past advice and learnings when facing a new task.
- `hindsight_store_experience`: Store an agent's experience after attempting a task. Useful for learning what works and what doesn't.
- `knowledge_forget_project`: Delete all knowledge chunks for a project.
- `knowledge_index`: Harvest and index a workspace project into searchable knowledge chunks. project: logical project name used for namespacing. root: optional absolute path to the 
- `knowledge_reindex`: Clear a project's knowledge index and re-harvest it.
- `knowledge_search`: Search knowledge chunks using hybrid semantic + keyword retrieval.
- `knowledge_stats`: Return knowledge index statistics, optionally filtered by project.
- `list_directory`: List directory contents with basic metadata.
- `memory_evaluate`: Change validation status: pending | verified | rejected.
- `memory_forget`: Delete a memory by ID.
- `memory_quantize`: Compress a memory to 'summary' or 'compressed' level.
- `memory_recall`: Recall relevant memories using hybrid semantic + keyword search.
- `memory_search`: Search memories using structured filters (tags, category, project, recency).
- `memory_stats`: Return memory statistics, optionally filtered by namespace.
- `memory_store`: Save a memory to long-term storage. category: general | decision | learning | context | error | pattern importance: 1-10 expires_in_days: optional TTL after whi
- `memory_update`: Update specific fields of an existing memory.
- `mikrotik_export_config`: Export the full RouterOS configuration (via SSH ``/export``). Args: router: router profile to use ('default' or 'router2'). Default 'default'. router_host: dyna
- `mikrotik_get_identity`: Read the router's system identity (name). Args: router: router profile to use ('default' or 'router2'). Default 'default'. router_host: dynamic override for rou
- `mikrotik_get_interfaces`: List network interfaces and their status. Args: router: router profile to use ('default' or 'router2'). Default 'default'. router_host: dynamic override for rou
- `mikrotik_get_ip_addresses`: List configured IP addresses on the router. Args: router: router profile to use ('default' or 'router2'). Default 'default'. router_host: dynamic override for r
- `mikrotik_get_system_resource`: Get RouterOS version, uptime, CPU/board/memory info. Args: router: router profile to use ('default' or 'router2'). Default 'default'. router_host: dynamic overr
- `mikrotik_ping`
- `mikrotik_run_rest`
- `mikrotik_ssh_command`
- `move_file`: Move or rename a file within allowed directories.
- `read_file`: Read a text file from an allowed directory.
- `run_shell`: Run a single whitelisted shell command inside an allowed working directory.
- `sanitize_string`: Sanitize a string by removing dangerous characters.
- `search_files`: Search files by glob pattern inside an allowed directory.
- `skill_delete`: Delete a skill by name and namespace.
- `skill_list`: List registered skills.
- `skill_load`: Load a single skill's full prompt template.
- `skill_load_registry`: Bulk-load skill definitions from a directory of YAML/JSON files.
- `skill_recall`: Recall skills matching the query using semantic + keyword search.
- `skill_register`: Register a reusable skill / procedural memory pattern.
- `skill_update`: Update specific fields of an existing skill.
- `telegram_get_updates`: Get recent updates (messages) for the configured bot.
- `telegram_send_message`: Send a text message to a Telegram chat.
- `validate_input`: Validate an input string against a schema (email, url, uuid, alphanumeric, safe_string).
- `vision_ocr`: Run OCR on an image file using Google Cloud Vision.
- `write_file`: Write text content to a file inside an allowed directory.

Measured by Glimind (https://glimind.com): MCP handshake + tool list only, no tool is ever called. CC BY 4.0.
