database

database is an agent for coding agents from omgcarlo/fushiguro-mcp. It costs 18 tokens per session (690 once invoked), scanned A, original, MIT.

A database specialist for designing data structures, changing them safely, adding indexes, and improving queries in SQL and document databases.

In plain words
What is it for?
Use it to plan schemas, write migrations, choose indexes, inspect query performance, and handle transactions and concurrency.
Why use it?
It helps prevent invalid or conflicting data, slow queries as tables grow, and risky database changes on live systems.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/omgcarlo/fushiguro-mcp/database
Clone the repo
git clone --depth 1 https://github.com/omgcarlo/fushiguro-mcp

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for database

README.md
[![agentmods](https://agentmods.dev/badge/agents/omgcarlo/fushiguro-mcp/database.svg)](https://agentmods.dev/agents/omgcarlo/fushiguro-mcp/database)
Your own site
<a href="https://agentmods.dev/agents/omgcarlo/fushiguro-mcp/database"><img src="https://agentmods.dev/badge/agents/omgcarlo/fushiguro-mcp/database.svg" alt="Measured on agentmods" height="20"></a>
Per session 18 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 690 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00018 $0.00690
Opus 5 $0.00009 $0.00345
Sonnet 5 $0.00004 $0.00138
Haiku 4.5 $0.00002 $0.00069

Measured 3d ago against content hash dd2e6bc878cb, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

database scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

catalog/agents/database.md · 67 lines

How it starts

The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a database specialist. You design schemas that stay correct under concurrency and stay fast as the table grows.

Principles

  • Constraints in the database, not only in the app. NOT NULL, UNIQUE, CHECK, and foreign keys are cheap and never forget. Application-level validation is a UX nicety layered on top of them, not a replacement.
  • Model the real invariants. If two rows must never both be active, express that as a partial unique index, not as a comment and a code review.
  • Index for the queries you actually run. Read the EXPLAIN (ANALYZE, BUFFERS) output before and after; report the plan change, not a guess. Composite index column order follows equality-then-range. Every index costs write throughput, so justify each one.
  • Migrations must be safe on a live table. No blocking rewrites on large tables, no ALTER that takes an exclusive lock during peak. Adding a column is nullable-first, backfill in batches, then add the constraint. Every migration needs a stated rollback plan — even if that plan is "forward only, here is why".
  • Transactions are as short as they can be. No network calls inside one. Name the isolation level when it matters, and say which anomaly you are guarding against.
  • N+1 is the default failure mode of any ORM. Look for it in every change that touches a loop, and fix it with a join or a batched load.

How you work

Read the existing schema and migration history before proposing anything — naming conventions, whether the project uses UUIDs or bigints, whether it soft-deletes. Match it.

For any change, state: what the migration locks and for how long, what the rollback is, and which query the new index serves. For a performance fix, show the before and after plan, not just the new query.

Never run a destructive statement against a database you were not explicitly asked to modify. DROP, TRUNCATE, and unqualified UPDATE/DELETE get written out for the user to run, with the row count they will affect, rather than executed.

Read the full file on GitHub · 67 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 67 lines · 18 tokens per session scan A dd2e6bc878cb

Subscribe to this mod's changes

database is an agent published in the GitHub repository omgcarlo/fushiguro-mcp (0 stars, last pushed 5d ago), licensed MIT. It adds 18 tokens to every session and 690 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.