mcp-server

A set of project rules for building MCP servers, including their tools, data resources, input schemas, responses, and safety checks. MCP is a standard way for an AI assistant to use external tools.

In plain words
What is it for?
It guides tool registration, schema design, success and failure responses, resource redaction, and controls such as allowing only approved single-statement database operations.
Why use it?
It encourages predictable tool behaviour, strict input validation, safe errors, and limits on potentially risky database queries.

Cursor rule for Cursor

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 rules/evtapps/ditto-mcp/mcp-server
Clone the repo
git clone --depth 1 https://github.com/evtapps/ditto-mcp

Made for: Cursor.

Per session 15 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 351 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 $0.00015 $0.00351
Opus 5 $0.00008 $0.00176
Sonnet 5 $0.00003 $0.00070
Haiku 4.5 $0.00002 $0.00035

Measured yesterday against content hash 7387811028d1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

mcp-server 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 yesterday.

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.

.cursor/rules/mcp-server.mdc · 31 lines

What it actually says

MCP Server Practices

Tools

  • Register tools in createServer() only; keep handlers small and deterministic.
  • inputSchema: define with Zod field shapes (e.g., { param: z.string() }) or a z.object({ ... }) shape assigned to a constant and passed as fields. Keep numeric constraints (int, positive, upper bounds) and url() where applicable.
  • Validate user input strictly (minLength, numeric constraints, URL where applicable).
  • Return { content: [...] } on success; { isError: true, content: [...] } on failure. Keep error content concise and actionable.
  • Error content should be concise, actionable, and safe (no secrets).

Resources

  • Use readable titles and descriptions. Keep resource bodies small and redacted if they include configuration. Only expose non-sensitive fields in resources like ditto://config.

Guardrails

  • Always enforce single-statement DQL and detect the first token (SELECT|INSERT|UPDATE|DELETE|EVICT). Strip comments before tokenizing.
  • Enforce operation allow-list from config. Default to SELECT only.
  • If configured, require the DQL to match at least one allow-list regex.

Configuration Access

  • Tool handlers do not read env directly. Resolve base URL, API key, and timeouts from the ServerConfig built in loadConfig().

Errors

  • Prefer returning an error content block over throwing; throwing should be reserved for truly unexpected conditions.
  • The Ditto client should convert failures into a Ditto-like envelope to keep tool handlers simple.
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. yesterday First seen · 31 lines · 15 tokens per session scan A 7387811028d1

Subscribe to this mod's changes

mcp-server is a cursor rule published in the GitHub repository evtapps/ditto-mcp (0 stars, last pushed 9mo ago), licensed MIT. It adds 15 tokens to every session and 351 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.