agents-md

A set of rules for keeping AGENTS.md as living documentation for coding agents. The file records project commands, conventions, architecture decisions, and unusual constraints.

In plain words
What is it for?
Use it to create or update AGENTS.md with build, test, lint, run, structure, naming, and decision information as the repository changes.
Why use it?
It gives future agents a consistent place to find practical project guidance without filling the human-focused README with agent-specific details.

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/jacoby6000/smithplates/agents-md
Clone the repo
git clone --depth 1 https://github.com/Jacoby6000/Smithplates

Made for: Cursor.

Per session 463 This file is loaded in full into every session.
When invoked 463 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 $0.00463 $0.00463
Opus 5 $0.00231 $0.00231
Sonnet 5 $0.00093 $0.00093
Haiku 4.5 $0.00046 $0.00046

Measured yesterday against content hash 5f43d7310b6e, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agents-md 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 yesterday.

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.

.cursor/rules/agents-md.mdc · 46 lines

What it actually says

AGENTS.md

Every repository should have an AGENTS.md at its root — the predictable place for agent-facing context that would clutter a README (build/test/lint commands, conventions, architecture decisions, gotchas). Keep it separate from README.md, which stays focused on human contributors. See https://agents.md/.

Maintain it as you work

Treat AGENTS.md as living documentation and update it regularly, without being asked, as the codebase grows:

  • Record decision points: when you make or discover a non-obvious choice (an approach taken, a constraint, a rejected alternative and why), capture it so future agents don't relitigate it. This complements in-code DESNOTEs — durable, cross-cutting decisions belong here.
  • Keep commands current: build, test, lint/format, and run commands an agent needs. Agents will execute listed checks, so they must be accurate.
  • Capture conventions and gotchas: project structure, naming, and anything strange or surprising about the environment.
  • If a repository has no AGENTS.md, create one when you begin substantive work.

Detailed reference docs (.ai-doc-reference/)

Keep AGENTS.md lean. For detailed information that is valuable but too large or specialized to keep in context at all times, write a dedicated document under .ai-doc-reference/ that agents fetch on demand. Use this for things like persistent CAD documents, schemas/contracts, deep design write-ups, runbooks, and data-model references.

  • Produce thorough, self-contained docs there as the codebase grows — favor detail over brevity, since these are loaded only when needed.
  • From AGENTS.md, link to each reference doc with a one-line description of what it contains and when to read it, so agents know what is available to pull in without loading it eagerly.

Monorepos

Use nested AGENTS.md files inside subprojects/packages. The file nearest the edited code wins; an explicit user instruction overrides any AGENTS.md.

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. yesterday First seen · 46 lines · 463 tokens per session scan A 5f43d7310b6e

Subscribe to this mod's changes

agents-md is a cursor rule published in the GitHub repository Jacoby6000/Smithplates (2 stars, last pushed 21d ago), licensed MIT. It adds 463 tokens to every session, about $0.0023 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.