Agent OS

Inject Standards

Deploy standards when you need them

What it does

The /inject-standards command deploys relevant standards into your current context. It can auto-suggest based on what you're working on, or inject specific standards explicitly.

Usage modes

Auto-suggest mode

Copy
/inject-standards

Analyzes your current conversation and suggests relevant standards based on the index descriptions.

Explicit mode

Copy
/inject-standards api                           # All standards in api/
/inject-standards api/response-format           # Single file
/inject-standards api/response-format api/auth  # Multiple files

Directly injects specified standards without suggestions.

How it works

1. Detect context

The command analyzes your current conversation to determine the best way to inject standards. Different contexts have different needs:

  • Conversation — You're chatting or implementing code. Standards should be loaded directly into context for immediate use. Content doesn't need to persist beyond this session.
  • Creating a skill — You're building a .claude/skills/ file. You may want to embed standards content (for a self-contained skill) or use file references (so the skill stays current when standards are updated).
  • Shaping/planning — You're in plan mode or building a spec. Similar choice: embed content for a point-in-time snapshot, or reference files for living documentation.

This detection happens automatically based on conversation signals—whether you're in plan mode, whether you've mentioned creating a skill, or what files you're working on.

2. Match and suggest

In auto-suggest mode, the command reads index.yml and matches descriptions against your context. It presents 2-5 relevant standards to confirm.

3. Inject based on scenario

Output format depends on scenario:

  • Conversation — Reads full standards content directly into context for immediate use.
  • Skill/Plan — Asks how you want standards included:
    • File references — Points to the standards files (e.g., @agent-os/standards/api/auth.md). The skill or plan stays current when standards are updated.
    • Copy content — Embeds the full standards text directly. Creates a self-contained document, but won't reflect future standards updates.

The index

The index.yml file enables matching without reading all standards:

Copy
api:
  response-format:
    description: API response envelope structure, status codes
  error-handling:
    description: Error code conventions, exception handling

database:
  migrations:
    description: Migration naming, column conventions

Good descriptions help the command suggest the right standards.

Tips

  • Run early — Inject standards at the start of a task, before implementation
  • Be specific — If you know which standards apply, use explicit mode
  • Check the index — If suggestions seem wrong, run /index-standards to rebuild

When in conversation mode, the command also surfaces related Skills from .claude/skills/ that you might want to use. It won't invoke them automatically—just mentions them for awareness.

Stay updated on Agent OS

Get notified of major new releases and updates to Agent OS.

Builder Methods

Training and community for developers and founders building with AI.

Created by Brian Casel (that's me) — multi-time founder, product designer, teacher, and the creator behind Builder Methods.

© 2026 CasJam Media, LLC / Builder Methods