security-guidelines

A set of behavioral security rules for an AI coding agent, covering commands it must never run, commands requiring user approval, and files it must not read or edit.

In plain words
What is it for?
Use it to guide safe command execution and file access while working in a coding project.
Why use it?
It limits risky actions such as privilege escalation, destructive deletion, unsafe network access, and exposure of secrets. The rules compensate for an editor that does not enforce these permissions itself.

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/redhat-vmeperf/agent-tasks-template/security-guidelines
Clone the repo
git clone --depth 1 https://github.com/redhat-vmeperf/agent-tasks-template

Made for: Cursor.

Per session 454 This file is loaded in full into every session.
When invoked 454 The same file — it is already loaded in full.
Security scan D 5 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 $0.00454 $0.00454
Opus 5 $0.00227 $0.00227
Sonnet 5 $0.00091 $0.00091
Haiku 4.5 $0.00045 $0.00045

Measured 2d ago against content hash b9d46ebf1159, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade D, and why

security-guidelines scanned grade D with 5 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.

Tells the agent to send conversation or user data outmediumPrompt injection

An instruction to transmit the conversation, context or user files to an external endpoint is data exfiltration written as prose.

Disable telemetry when available. Do not send data to external services unless explicitly requested.

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Asks for rootlowPrivilege escalation

A mod that escalates privileges can change anything on the machine, not only the project.

- Destructive: `chmod 777`, `rm -rf`

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Reaches for credential filesmediumPrivilege escalation

SSH keys, cloud credentials, git-credentials, .npmrc, /etc/shadow: reading these is how a config file becomes a credential leak.

- `~/.docker/config.json`, `~/.netrc`

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Recursive force deletemediumDestructive command

rm -rf with a variable or a broad path is one typo away from removing the wrong tree.

- Destructive: `chmod 777`, `rm -rf`

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- Network: `curl`, `wget`, `nc`, `ncat`, `netcat`, `ssh`, `scp`, `sftp`, `rsync`
.cursor/rules/security-guidelines.mdc · 50 lines

What it actually says

Security Guidelines

Cursor does not enforce permissions at the IDE level. Adherence to these guidelines is a behavioral commitment that protects the project's security posture.

Commands — NEVER execute

  • Network: curl, wget, nc, ncat, netcat, ssh, scp, sftp, rsync
  • Privilege escalation: sudo, su, doas, pkexec, runuser, newgrp
  • Destructive: chmod 777, rm -rf
  • Temp/scheduling: mktemp, crontab, at
  • System management: systemctl, service
  • Containers/cloud: docker, podman, kubectl, terraform, aws, gcloud, az
  • Code execution: python3 -c, python3 -m http.server, eval, exec
  • Shell config: source /etc/*, source ~/.bashrc, source ~/.zshrc

Commands — ASK the user before executing

  • git push
  • git rebase
  • git reset
  • git merge

Files — NEVER read

  • .env, .env.*
  • secrets/**
  • *.pem, *.key, *.p12, *.pfx
  • ~/.ssh/**, ~/.aws/**, ~/.azure/**, ~/.kube/**, ~/.gnupg/**
  • ~/.docker/config.json, ~/.netrc
  • /etc/passwd, /etc/shadow, /etc/sudoers

Files — NEVER edit

  • .git/**
  • /etc/**, /usr/**, /bin/**
  • ~/.bashrc, ~/.zshrc, ~/.profile, ~/.ssh/**

Web fetching — DISABLED

Do not fetch external URLs.

Telemetry

Disable telemetry when available. Do not send data to external services unless explicitly requested.

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 · 50 lines · 454 tokens per session scan D b9d46ebf1159

Subscribe to this mod's changes

security-guidelines is a cursor rule published in the GitHub repository redhat-vmeperf/agent-tasks-template (3 stars, last pushed 4mo ago), licensed Unlicense. It adds 454 tokens to every session, about $0.0023 per session on Opus 5. A static security scan graded it D with 5 findings (tells the agent to send conversation or user data out, asks for root, reaches for credential files). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.