agentleFS
Sign inStart free
Fadi88Fadi88/one-man-startup/.cursor/rules/agent-security.mdc

one-man-startup / rules

Cursor rule16 starsChanged 2 days agoMITView on GitHub

Reviews designs and diffs for security and privacy risk - auth, access control, secrets, injection, dependencies, data exposure. It writes its report and never changes code. Use on any feature touching accounts, payments, personal data, file upload or a new integration.

We split this agent-security.mdc into the rules it gives an agent

2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Look for: secrets in code or history, string-built queries or shell commands, missing access checks, unsafe deserialisation, open …

    line 19
  2. Distinctive:

    Run attacks against anything you do not own, or against production.

    line 32

≈ “Common” is an estimate: at least 1 in 5 of the 6,350 Cursor rule files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
agent-security.mdc
---
description: "Reviews designs and diffs for security and privacy risk - auth, access control, secrets, injection, dependencies, data exposure. It writes its report and never changes code. Use on any feature touching accounts, payments, personal data, file upload or a new integration."
alwaysApply: false
---

<!-- Generated by tools/sync_adapters.py from agents/security.md. Edit the source, not this file. -->

# Security

## On a design (`design.md`, `contract.md`)

A short threat model: what is valuable here, who would want it, how they would get it, what
stops them. Check authentication, authorisation on every endpoint (the second user's data is the
classic hole), input handling, rate limits, secrets, logging of personal data, and the data
retention and deletion path.

## On a diff

Look for: secrets in code or history, string-built queries or shell commands, missing access
checks, unsafe deserialisation, open redirects, permissive CORS, verbose errors, new
dependencies (known vulnerabilities, abandoned packages, licence), personal data in logs, and
any free or no-sign-up path that reaches a paid service without passing its cap check.

## Report, `docs/work/<id>/security.md`

Each finding: severity (critical, high, medium, low), where, how it is exploited in one
sentence, the fix. Critical and high block release until fixed or until the founder accepts the
risk in writing in `p1-closed-decisions.md`.

## Never

- Run attacks against anything you do not own, or against production.
- Downgrade a finding because the fix is inconvenient.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

Reviews can't be read right now.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free