"Understanding File Usage" matching MCP servers:

An MCP server that provides persistent memory for database schemas by storing table structures and metadata in a local SQLite file. It enables LLMs to save, retrieve, and manage database snapshots to maintain a structured understanding of database architectures.

Enables AI coding assistants to deploy and manage compatible web apps, isolated Postgres databases, and project file storage, while inspecting build logs, runtime errors, and usage through an OAuth-authenticated remote MCP connection.

Provides AI assistants with persistent graph database memory using Neo4j, enabling task management, relationship understanding, semantic search with embeddings, file indexing, and multi-agent coordination through the Model Context Protocol.

Enables minimal interaction with Supabase databases through 4 essential tools for data querying, mutations, file storage, and user authentication. Designed for 70% less context usage than standard implementations with auto-truncated results and simplified parameters.

An MCP server that gives LLM clients access to a SQL Server database and the local filesystem, including file operations and image handling.

Manages LinkedIn prospecting CSV files with deduplication and search, reducing LLM token usage by offloading file operations.

Enables natural language analysis of Azure usage data from CSV files, providing cost summaries, visualizations, and insights about service and regional spending patterns.

A read-only DuckDB MCP server offering context-efficient analytics tools (list_datasets, describe_table, profile_column, explain, query) with a semantic layer for business rules, security guards, and disclosed truncation to help LLMs produce correct answers while minimizing token usage.

Lets an agent query JSON files too large for its context: JSONPath with aggregation pipes, schema, field profiles, samples, identity-keyed diff, CSV and spreadsheet export, and checks that gate a file. Int64 IDs and decimals keep their exact digits, and zstd, Base64-Zstd and PostgreSQL bytea payloads open directly.

Provides read-only visibility into Snowflake warehouse cost and query performance, including credit usage by warehouse, expensive queries, and heuristic right-sizing recommendations.

A Model Context Protocol server providing tools for DB queries, API calls, file I/O, and text transformations, enabling AI agents like Claude to perform real-world actions.

Marker-indexed SQL cache MCP server that indexes in-file @scry.\* markers into a SQLite database, enabling agents to query structured project knowledge via read-only SQL without LLM reasoning.

Enables AI assistants to interact with Hadoop Hue for executing SQL queries using Hive, SparkSQL, or Impala and managing HDFS files. It supports directory browsing, file transfers, and exporting query results to CSV through the Model Context Protocol.

Enables an AI assistant to read form structures and submissions from a Coletum account (read-only) and to produce Excel/CSV exports, Coletum-style or user-template PDFs, and photographic reports on request. It also stores preferences such as logo and company name and reports where each generated file was saved.

Query only the data you need from your JSON file and keep your context clean

Enables managing PocketBase instances through intent-first tools for schema inspection, record CRUD, bulk operations, authentication, file handling, and server monitoring, with optional destructive operations gated behind explicit opt-in.

Exposes 3 meta-tools to efficiently search and execute 174 backend CRUD tools, reducing context token usage by ~98%.

A read-only MCP server that exposes SQL database access to LLMs, supporting multiple database types, compact columnar results, pagination, and file export.

Read-only MCP server for searching your local Retrace screen-history database, with tools for full-text search, segment listing, frame details, app usage, and tags.

Enables Claude Code to work with personal data that stays on the local machine, running SQL, Python, and file/schema inspection in a sandboxed compute-to-data setup that returns only aggregates, schemas, synthetic fixtures, and pseudonymized results. Raw records, file names, and protected values never reach the Anthropic API, so Auto mode can stay hands-off.

Turns any JSON file (catalogue, FAQ, price list) into an MCP server with three read-only tools: list with paging, keyword search across chosen fields, and get by id. Runs over stdio for Claude Desktop, Claude Code and Cursor.

Enables file operations, document metadata extraction, folder analysis, local-LLM queries, and SQLite CRUD as MCP tools, individually or composed into pipelines via run_pipeline.

Enables Claude Code and Gemini CLI to save, search, retrieve, and browse recent voice-derived notes and tags stored in SQLite over stdio, while deliberately excluding deletion and arbitrary file reads.

Enables an engineering assistant to persist and query local CAD review context—named features, notes, tags, and revision-bound decisions with their requirements and evidence—in SQLite by artifact, entity, text, or tag, marking records stale against a caller-supplied revision. It never opens the CAD file, so it works on a laptop with no CAD software or other server running.

MCP server for tabular data retrieval that indexes local CSV, Excel, Parquet, and JSONL files once and answers questions via column profiles, filtered rows, server-side aggregations, and joins, drastically reducing token usage for large datasets.

Context-efficient MCP server that executes OpenSearch DSL queries and returns synthesized output instead of raw JSON, reducing token usage by up to 90%.

A TypeScript MCP server for interacting with CDISC Define-XML 2.0 and 2.1 files, enabling users to load, search, and retrieve clinical data metadata including datasets, variables, and codelists. It supports local file processing with document caching and provides responses in both JSON and Markdown formats.

A unified MCP server for multi-stand infrastructure testing, enabling REST API calls, PostgreSQL queries, and OpenSearch searches through a single config file.

A comprehensive MCP server for Dokploy that provides 14 action-based tools covering the full DevOps lifecycle with minimal token usage. It enables AI assistants to manage projects, applications, databases, domains, servers, and more through a unified interface.

Enables an AI client to search Australian federal and South Australian government open data catalogues, inspect a dataset file's schema and sample rows, then run read-only SQL over the full file in a sandboxed DuckDB. Every answer carries the source URL, publisher and licence, with extra tools for ABN checksum validation and Australian financial-year dates.