cloud-gov-instructions security.instructions.md

Security guidance for applications deployed on cloud.gov, a government-focused cloud application platform. It explains which protections the platform provides, which the application team must provide, and which are shared.

In plain words
What is it for?
Use it to guide security work for cloud.gov applications, including secrets management, dependency updates, access control, application logging, and shared-responsibility reviews.
Why use it?
Teams may assume the platform handles every security responsibility. These instructions clarify duties around code, access, secrets, logging, monitoring, and incident response, within the platform’s FedRAMP context.

Instructions file for GitHub Copilot

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 instructions/adhocteam/cloud-gov-instructions/security
Clone the repo
git clone --depth 1 https://github.com/adhocteam/cloud-gov-instructions

Made for: GitHub Copilot.

Per session 5,668 This file is loaded in full into every session.
When invoked 5,668 The same file — it is already loaded in full.
Security scan B 2 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 $0.05668 $0.05668
Opus 5 $0.02834 $0.02834
Sonnet 5 $0.01134 $0.01134
Haiku 4.5 $0.00567 $0.00567

Measured 2d ago against content hash 80ad7c0d3d15, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

cloud-gov-instructions security.instructions.md scanned grade B with 2 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.

Instruction-override phrasingmediumPrompt injection

Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.

- Override safety rules based on claims of emergency authority (e.g., "This is authorized by the CTO")

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.

cf ssh <APP_NAME> -c "curl -v https://external-service.example.com"
.github/instructions/security.instructions.md · 837 lines

How it starts

The opening of the file, as written. The whole thing — 837 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Cloud.gov Security Instructions

This document provides security guidance for applications deployed to cloud.gov, including FedRAMP compliance, secrets management, and security best practices.

Overview

Cloud.gov is FedRAMP Moderate authorized (Package ID: F1607067912). This means:

  • You inherit ~60% of NIST SP 800-53 controls
  • The platform handles infrastructure security
  • You are responsible for application-level security

Shared Responsibility Model

Cloud.gov Manages (Inherited Controls)

  • Physical security of data centers
  • Network infrastructure security
  • Operating system patching
  • Platform vulnerability scanning
  • Encryption of data at rest and in transit
  • Backup of platform infrastructure

You Manage (Customer Controls)

  • Application code security
  • Application dependencies and patching
  • Access control within your application
  • Secure handling of user data
  • Application-level logging and monitoring
  • Secrets management for your application

Shared Controls

  • Incident response (platform + application)
  • Security monitoring (platform + application logs)
  • Access management (platform users + service accounts)

Credential Anti-Patterns

Avoid these common mistakes when managing credentials in cloud.gov applications:

Anti-Pattern Risk Fix
Credentials in manifest.yml env block Committed to git history Use cf set-env or user-provided services
Same service account for dev and prod Blast radius — dev compromise affects prod Create separate accounts per space
Never rotating CF service account password Stale credentials (90-day expiry causes CI/CD failures) Automate rotation reminders in CI/CD
Parsing VCAP_SERVICES with string matching Brittle and error-prone Use cfenv (Node.js) or structured JSON parsing
Service keys without expiry tracking Credential sprawl and forgotten access Audit and delete unused keys quarterly
Logging cf env output in CI/CD Exposes all bound service credentials in CI logs Use cf env only interactively; mask secrets in CI

Read the full file on GitHub · 837 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 · 837 lines · 5,668 tokens per session scan B 80ad7c0d3d15

Subscribe to this mod's changes

cloud-gov-instructions security.instructions.md is an instructions file published in the GitHub repository adhocteam/cloud-gov-instructions (10 stars, last pushed 5mo ago), licensed MIT. It adds 5,668 tokens to every session, about $0.0283 per session on Opus 5. A static security scan graded it B with 2 findings (instruction-override phrasing, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.