Verified

Firebird MCP Server

Firebird SQL database query analysis and performance tuning for AI assistants.

Data last scanned today · Reviewed today

5022 tools★ 9PascalNOASSERTION

Overview

Firebird MCP Server allows AI assistants to interface directly with Firebird database instances ranging from version 2.5 to 5.0. By querying monitoring tables and access plans, the server enables users to identify performance bottlenecks, diagnose missing indexes, and pinpoint transactions obstructing garbage collection. Every diagnostic interaction provides a structured finding, a recommended SQL fix, and verification steps. The server operates in a read-only mode, ensuring that the AI provides analysis without altering your database directly.

Our verdict

This server carries a high risk level (30/100) and should be evaluated with caution. The project is community-maintained rather than vendor-backed, and the NOASSERTION license provides no clear legal clarity on usage terms. Potential users should carefully review the implementation as the authentication methods have not yet been editorially verified.

Supports Firebird versions 2.5 through 5.0
Provides actionable SQL fixes for database performance issues
Zero direct dependencies reduce supply chain risk
Standardized JSON-RPC 2.0 communication over stdio

NOASSERTION license status lacks legal clarity
Community-maintained project without official vendor support
Single contributor presents potential bus-factor risk

Tools

Tool	Description	Risk
fb_info	Engine version + detected capabilities (JSON).	low
fb_list_tables	List user tables.	low
fb_generate_documentation	Markdown docs (columns, PK, indexes, triggers with the event they fire on) for one table, or the whole database.	low
fb_analyze_query	Access-plan analysis (NATURAL-scan + external-SORT detection), the engine's explained plan, and measured cost.	low
fb_suggest_indexes	Index advice for one query, including refreshing stale statistics.	low
fb_suggest_index_drops	Flags duplicate, redundant-prefix, inactive, or low-selectivity indexes.	low
fb_audit_table	Schema-health audit including missing PK, over-indexing, and stale statistics.	low
fb_evaluate_goal	Deterministic goal check (drives the optimization loop).	low
fb_monitor_transactions	Monitor database transactions.	low
fb_sample_activity	Sample database activity.	low

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›Which Firebird versions does this server support?

The server supports Firebird versions 2.5, 3.0, 4.0, and 5.0, with capability detection performed automatically at runtime.

›Does the server require specific database permissions?

The server attaches to your database in read-only mode to perform its analysis and does not write to your database.

›Do I need to provide a client library for the server?

Yes, you must provide your own fbclient.dll, as the server zip does not include it by default.

›Is this server free to use?

The software is free for use on your own databases at any scale without keys or expiry; a license is only required if you pass the software on to another party.

›How does the server perform backups?

The server runs gbak in the background and subsequently restores the backup into a scratch database to verify the backup is functional.