project

A set of project rules for agents working in Cursor, an editor for writing code. The rules cover how to inspect files, plan sensitive changes, follow repository conventions, and verify work.

In plain words
What is it for?
Use it to guide edits, database or authentication changes, testing, linting, type checks, builds, dependency use, generated files, and security-sensitive work.
Why use it?
It keeps changes small and consistent while preventing unapproved features, unsafe database or deployment actions, and committed secrets. It also tells agents which checks to run before finishing.

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/francescogabrieli/glyph/project
Clone the repo
git clone --depth 1 https://github.com/francescogabrieli/glyph

Made for: Cursor.

Per session 740 This file is loaded in full into every session.
When invoked 740 The same file — it is already loaded in full.
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 $0.00740 $0.00740
Opus 5 $0.00370 $0.00370
Sonnet 5 $0.00148 $0.00148
Haiku 4.5 $0.00074 $0.00074

Measured 2d ago against content hash def5b0252c8e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

project 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 2d 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.

benchmarks/cursor-rules/.cursor/rules/project.mdc · 73 lines

How it starts

The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cursor Project Rules

Read relevant files before editing. Create a short plan before changing database schema, authentication, or shared components. Keep changes small and reviewable. Prefer simple explicit code. Follow existing repository conventions.

  • Before marking work done, run the relevant tests.
  • Run lint before done.
  • Run type checks before done.
  • Do not add unrequested features.
  • Avoid large refactors and broad rewrites.
  • Report what changed and report what was verified.
  • If verification was skipped, explain unrun tests.

Commands:

pnpm install
pnpm dev
pnpm test
pnpm lint
pnpm build

Security:

Never commit secrets, credentials, API keys, tokens, or private keys. Ask for confirmation before destructive operations, schema changes, unsafe migrations, and security-sensitive changes. Do not deploy without approval.

Do not edit generated files manually. Use shared database helpers rather than raw connections.

Repository notes

Cursor rules often mix direct behavioral instructions with notes about editor behavior, generated files, and project conventions. This project keeps reusable data access helpers in a shared package so services do not open ad hoc database connections. That guidance is repository-specific and useful, but it is not a universal Glyph semantic unit unless a custom rule is supplied.

The benchmark keeps frontmatter, checklist-like rules, command blocks, safety notes, and prose in one file because real .mdc files commonly combine those forms. The structural parser should keep line numbers and section context so inspect output is actionable.

Cursor workflow

Always read relevant files before editing. Create a short plan before non-trivial edits. Make the smallest safe change. Follow existing style, prefer existing patterns, and keep changes small and reviewable.

Use shared database helpers rather than raw connections. Use existing shared helpers before introducing new abstractions. Avoid unrelated packages. Respect package boundaries and ask before touching another workspace.

Read the full file on GitHub · 73 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. 2d ago First seen · 73 lines · 740 tokens per session scan A def5b0252c8e

Subscribe to this mod's changes

project is a cursor rule published in the GitHub repository francescogabrieli/glyph (7 stars, last pushed 1mo ago), licensed MIT. It adds 740 tokens to every session, about $0.0037 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.