agent-plugins AGENTS.md

A guide to the AWS agent-plugin project, where plugins bundle instructions, external tool connections, and agent configurations. It explains how plugins, skills, tool servers, and marketplaces differ.

In plain words
What is it for?
Use it when developing, documenting, or organizing AWS plugins, skills, external tool connections, or marketplace entries.
Why use it?
It helps an agent understand which extension mechanism fits a task and how automatic skill selection works. This reduces confusion when adding or using AWS-focused agent capabilities.

Instructions file for CodexOpenCode

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/awslabs/agent-plugins/agents-md
Clone the repo
git clone --depth 1 https://github.com/awslabs/agent-plugins

Made for: Codex, OpenCode.

Per session 2,127 This file is loaded in full into every session.
When invoked 2,127 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 $0.02127 $0.02127
Opus 5 $0.01064 $0.01064
Sonnet 5 $0.00425 $0.00425
Haiku 4.5 $0.00213 $0.00213

Measured 2d ago against content hash 5d88f864deb8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agent-plugins 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 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.

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.

AGENTS.md · 205 lines

How it starts

The opening of the file, as written. The whole thing — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agent Plugins for AWS

Note: CLAUDE.md is a symlink to this file. Only edit AGENTS.md — changes apply to both automatically.

TL;DR Pitch

This repository supports plugins - bundles of skills, MCP servers, and agent configurations that extend capabilities. The awslabs/agent-plugins marketplace includes plugins like deploy-on-aws (architecture recommendations, cost estimates, and working IaC), amazon-location-service (maps, geocoding, routing, and geospatial features), and databases-on-aws (database guidance for the AWS portfolio, starting with Aurora DSQL).

Core Concepts

Plugins vs Skills vs MCP Servers

Concept What It Is Example
Plugin A distributable bundle (skills + MCP servers + agents) deploy-on-aws
Skill Instructions that auto-trigger based on user intent (YAML frontmatter description) "deploy to AWS" triggers the deploy skill
MCP Server External tool integration via Model Context Protocol awspricing for cost estimates
Marketplace Registry of plugins users can install awslabs/agent-plugins

Key Design Decision: Skills Auto-Trigger

Skills are NOT slash commands. The agent determines when to use a skill based on the description field in YAML frontmatter. If user says "host this on AWS", the agent matches that intent to the deploy skill's description and invokes it.

Directory Structure

agent-plugins/
├── .claude-plugin/
│   └── marketplace.json          # Marketplace registry
├── .github/
│   ├── workflows/                # CI (build, lint, security, etc.)
│   ├── ISSUE_TEMPLATE/
│   └── ...
├── docs/                         # Role-specific guides
│   ├── DESIGN_GUIDELINES.md      # Plugin design best practices
│   ├── DEVELOPMENT_GUIDE.md      # Contributor setup and workflow
│   ├── MAINTAINERS_GUIDE.md      # Reviewer/maintainer processes
│   └── TROUBLESHOOTING.md        # Plugin troubleshooting
├── plugins/
│   ├── deploy-on-aws/
│   │   ├── .claude-plugin/
│   │   │   └── plugin.json       # Plugin manifest
│   │   ├── .mcp.json             # MCP server definitions
│   │   └── skills/
│   │       └── deploy/
│   │           ├── SKILL.md     # Main skill (auto-triggers)
│   │           └── references/
│   │               ├── defaults.md
│   │               ├── cost-estimation.md
│   │               └── security.md
│   ├── amazon-location-service/
│   │   ├── .claude-plugin/
│   │   │   └── plugin.json
│   │   ├── .mcp.json
│   │   └── skills/
│   │       └── amazon-location-service/
│   │           ├── SKILL.md
│   │           └── references/
│   ├── databases-on-aws/
│   │   ├── .claude-plugin/
│   │   │   └── plugin.json
│   │   ├── .mcp.json
│   │   ├── hooks/
│   │   │   └── hooks.json
│   │   ├── scripts/
│   │   └── skills/
│   │       └── dsql/
│   │           ├── SKILL.md
│   │           ├── mcp/
│   │           └── references/
├── schemas/                      # JSON schemas for manifests
│   ├── marketplace.schema.json
│   ├── plugin.schema.json
│   ├── mcp.schema.json
│   └── skill-frontmatter.schema.json
├── tools/                        # Lint, validation, and eval scripts
│   ├── evals/                    # Eval suites for plugins, kept separate from plugin clients
│   │   └── databases-on-aws/
│   │       ├── README.md
│   │       └── dsql/
│   │           ├── evals.json
│   │           ├── trigger_evals.json
│   │           ├── query_explainability_evals.json
│   │           ├── safe_query_evals.json
│   │           └── scripts/
│   │               └── run_functional_evals.py
│   ├── validate-cross-refs.cjs
│   └── ...
├── mise.toml                     # Tool versions and tasks
├── dprint.json
├── .markdownlint-cli2.yaml
├── .pre-commit-config.yaml
└── README.md

Read the full file on GitHub · 205 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 · 205 lines · 2,127 tokens per session scan A 5d88f864deb8

Subscribe to this mod's changes

agent-plugins AGENTS.md is an instructions file published in the GitHub repository awslabs/agent-plugins (876 stars, last pushed 6d ago), licensed Apache-2.0. It adds 2,127 tokens to every session, about $0.0106 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 2026-08-30.