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/kindrat86/agentmail/claude-mdgit clone --depth 1 https://github.com/kindrat86/agentmailWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/kindrat86/agentmail/claude-md)<a href="https://agentmods.dev/instructions/kindrat86/agentmail/claude-md"><img src="https://agentmods.dev/badge/instructions/kindrat86/agentmail/claude-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00658 | $0.00658 |
| Opus 5 | $0.00329 | $0.00329 |
| Sonnet 5 | $0.00132 | $0.00132 |
| Haiku 4.5 | $0.00066 | $0.00066 |
Grade A, and why
agentmail CLAUDE.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 3d 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.
What it actually says
sanctionsai.dev — граблі та правила
Деплой
- Fly.io, app =
agentmail-api:flyctl deploy(без нових secrets) - fly.toml: autostop=suspend — критичний фікс, який РЕВЕРТНЕТЬСЯ, якщо деплоїти зі старим fly.toml. Перед деплоєм перевір, що suspend на місці
Критичні граблі
- api.py = 836KB, всі manifests віддаються inline з нього — НІЯКОЇ хірургії api.py; статичний контент (glossary /learn/) — окремими файлами під served-префіксом
- Stripe webhook (fix 2026-08-15, v521): stripe>=9
construct_eventповертає StripeObject (без.get()) — handlers отримують plain dict черезjson.loads(payload)після верифікації підпису. Verified-but-ignored події = 200 (не 400), інакше Stripe ретраїть днями і спамить failure-емейли._stripe_webhookтепер логить кожну подію. - Compliance-контент: тільки VETTED verbatim визначення + disclaimers; verbatim gate обовʼязковий, нічого не вигадувати
- Dashboard фільтрує verify_* пінги; реальних відвідувачів історично ~0 — не інтерпретуй пінги як трафік
- Owner-задача (не робити самому): ротація ключів sanctionsai
- api.py.bak файли в репо — бекапи, не видаляти
Нові роути (DotCom Secrets audit 2026-07-26)
/playbook.pdf— лід-магніт (PDF, генеруєтьсяscripts/build_playbook.py)./startобіцяв його з launch; welcome-лист тепер його доставляє./checkout/dev/bump— order-bump interstitial (Ch14). МАЄ матчиситись ДО префікс-роуту/checkout/, інакше префікс зловитьdev/bumpяк план і 500не на відсутньому STRIPE ключі./upgrade— upsell Dev→Pro (Ch15), лінкається з post-purchase листа.- Tripwire
/tripwire— CTA тепер env-driven:STRIPE_PAYMENT_LINK_TRIPWIRE→ реальна Stripe кнопка, інакше робочий fallback (playbook + ручний invoice). /subscribeбільше не брешеemail_sent:true— повертає реальний результат +email_configured.- Owner-only (див.
OWNER_ACTIONS_DOTCOM_2026-07-26.md):RESEND_API_KEY,STRIPE_PAYMENT_LINK_TRIPWIRE,STRIPE_PRICE_BUMP_AUDIT.
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.
- 3d ago First seen · 23 lines · 658 tokens per session scan A fd1f4aacd0e3
agentmail CLAUDE.md is an instructions file published in the GitHub repository kindrat86/agentmail (1 stars, last pushed 3d ago), licensed MIT. It adds 658 tokens to every session, about $0.0033 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
mcp-custos CLAUDE.md
Instructions for an0malous/mcp-custos, covering claude.md — agent guide, project context, spec flow (human-approved gates), conventions and commands.
openchronicle-mcp CLAUDE.md
Claude Code instructions for CSOAI-ORG/openchronicle-mcp, covering claude.md, project status file, project-specific notes, current sprint and build and development.
FedRAMP20xMCP copilot-instructions.md
Copilot instructions for KevinRabun/FedRAMP20xMCP, covering copilot instructions for fedramp 20x mcp server, 1. machine execution contract (read first — non‑negotiable), 2. priority hierarchy (explicit goal ordering), 3. git flow workflow (mandatory) and ensure main is up to date.
FedRAMP20xMCP codacy.instructions.md
Configuration for AI behavior when interacting with Codacy's MCP Server.
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).
ai-toolkit AGENTS.md
AGENTS.md instructions for pipefy/ai-toolkit, covering repository guidelines, documentation map, project structure, import namespace migration: pipefysdk → pipefy and src/pipefysdk/init.py (transitional shim).