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.
npx agentmods add instructions/smarterweather/developer/agents-mdgit clone --depth 1 https://github.com/smarterweather/developerWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00686 | $0.00686 |
| Opus 5 | $0.00343 | $0.00343 |
| Sonnet 5 | $0.00137 | $0.00137 |
| Haiku 4.5 | $0.00069 | $0.00069 |
Grade A, and why
developer 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.
How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Smarter Weather developer platform
Guidance for AI agents integrating with Smarter Weather.
Surfaces
| Surface | Endpoint | Auth |
|---|---|---|
| REST API | https://api.smarterweather.com |
Authorization: Bearer sw_live_* / sw_test_* |
| Weather MCP | https://mcp.smarterweather.com |
OAuth 2.1 + PKCE, or API key Bearer |
| Onboarding MCP | https://mcp.developers.smarterweather.com |
Anonymous discovery tools; OAuth for account tools |
Canonical machine-readable contracts:
- OpenAPI:
openapi.yaml - Weather MCP tool descriptors:
descriptors/mcp-tools.json - Agent Plugin (skills + MCP config):
plugins/smarterweather/ - Agent doc index:
llms.txt
Preferred onboarding path
- Connect to the onboarding MCP with no credentials.
- Call
get_plans/get_documentation/sign_up.sign_upreturns a Clerk signup URL; a human completes account creation in the browser (no credit card for the free tier). - Complete Clerk OAuth when the host prompts.
- Call
create_api_key(idempotent) andconfigure_mcp. - Use the returned key against the weather MCP or REST API.
stdio bridges (local clients that cannot speak Streamable HTTP):
npx -y @smarterweather/mcp-onboarding
npx -y @smarterweather/mcp-weather
Do not append @preview — latest is the GA bridge. Set
SMARTERWEATHER_API_KEY in the process environment for headless weather
calls; set SMARTERWEATHER_ONBOARDING_AUTH=required to force onboarding
OAuth for account-scoped tools. Authenticated onboarding uses a
pre-registered public PKCE Clerk client (DCR off); ensure port 3334 is
free for the loopback callback. Prefer the stdio bridge over Cursor's
native url OAuth for gated tools (Clerk + cursor:// is broken).
Auth and errors
- Prefer
Authorization: Bearer <key>(notX-API-Key). - Rate-limit headers follow the
RateLimit-*family (see docs). - Errors use RFC 7807 problem details.
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.
- yesterday First seen · 67 lines · 686 tokens per session scan A 769ceef02696
developer AGENTS.md is an instructions file published in the GitHub repository smarterweather/developer (1 stars, last pushed 5d ago), licensed MIT. It adds 686 tokens to every session, about $0.0034 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.
Other instructions, from other repositories
tray-api-ai-plugin AGENTS.md
Instructions for tray-tecnologia/tray-api-ai-plugin, covering tray api plugin, regras obrigatórias para toda sessão, autenticação, url e formato and rate limit.
tray-api-ai-plugin CLAUDE.md
Instructions for tray-tecnologia/tray-api-ai-plugin, covering tray api plugin — contexto para claude, regras obrigatórias, bloco mandatory e lint de skills, validação local and busca em docs.
tray-api-ai-plugin copilot-instructions.md
Instructions for tray-tecnologia/tray-api-ai-plugin, covering tray api plugin, regras obrigatórias para toda integração tray, autenticação oauth 2.0, formato de requisições and rate limit.
substack-publisher-mcp CLAUDE.md
Instructions for dkships/substack-publisher-mcp, covering substack publisher mcp, commands, environment variables, conventions and definition of done.
builtwith-api CLAUDE.md
Instructions for zcaceres/builtwith-api, covering claude.md — builtwith-api, project overview, tech stack, key conventions and commands.
substack-publisher-mcp AGENTS.md
Instructions for dkships/substack-publisher-mcp, covering substack publisher mcp — agent instructions, what this is, tech stack, scope and boundaries and working rules.