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 skills add fmind/dot --skill draft-mailgit clone --depth 1 https://github.com/fmind/dotWrote 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/skills/fmind/dot/draft-mail)<a href="https://agentmods.dev/skills/fmind/dot/draft-mail"><img src="https://agentmods.dev/badge/skills/fmind/dot/draft-mail/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/fmind/dot/draft-mail"><img src="https://agentmods.dev/badge/skills/fmind/dot/draft-mail.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00017 | $0.00990 |
| Opus 5.5 | $0.00007 | $0.00396 |
| Sonnet 5.5 | $0.00003 | $0.00198 |
| Haiku 4.5 | $0.00002 | $0.00099 |
Grade A, and why
draft-mail 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 11d 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.
How it starts
The opening of the file, as written. The whole thing — 55 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Draft Mail
Write an email or reply as a Gmail draft the user reviews and sends from Gmail. This skill never sends: a request to "send" still produces a draft plus a handoff, unless the user explicitly approves sending this exact reviewed draft in the session. gws owns authentication, MIME, and request mechanics.
Workflow
-
Resolve the context: confirm the account with
gws auth status. For a reply, find the message and read the thread with bounded queries; extract the asks, deadlines, attachments mentioned, and every recipient.gws gmail users messages list --params '{"userId":"me","q":"from:[email protected] newer_than:30d","maxResults":5}' --format json gws gmail +read --id MESSAGE_ID --headers --format json -
Match language and register: write in the language of the thread, or of the recipient's last message to the user; when none exists, ask rather than guess. Read up to three of the user's sent messages to the same person (
q: "to:<address> in:sent") to mirror greeting, sign-off, and formality. In French, keep the thread'stuorvous(defaultvousfor clients and first contact), and French typography (Bonjour Alice,, a space before?and!when the user's mail uses one). -
Write in the user's voice: short paragraphs, the answer or ask first, concrete dates with weekday and timezone, one clear next step. Never invent commitments, prices, availability, attachments, or facts; mark anything the user must decide as a question in the handoff, not in the draft.
-
Create the draft from a body file in the scratchpad, plain text by default. Use
+reply(or+reply-allwhen every recipient should stay) for threads so headers and threading stay correct; both quote the original message automatically, so the body file holds only the new text. Preview with--draft --dry-run, then rerun without--dry-run; never remove--draft.gws gmail +reply --message-id MESSAGE_ID --body "$(cat "$SCRATCH/body.txt")" --draft --dry-run gws gmail +reply --message-id MESSAGE_ID --body "$(cat "$SCRATCH/body.txt")" --draft gws gmail +send --to [email protected] --subject 'Subject' --body "$(cat "$SCRATCH/body.txt")" --draft
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.
- 11d ago Changed c717af1673f8
- 12d ago First seen · 55 lines · 17 tokens per session scan A f56f3f3b16d2
draft-mail is a skill published in the GitHub repository fmind/dot (11 stars, last pushed yesterday), licensed MIT. It adds 17 tokens to every session and 990 once invoked, about $0.0001 per session on Opus 5.5. 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 2026-10-05.
Other skills, from other repositories
google-workspace-cli
Use gws across Workspace APIs with schema-first calls, profile isolation, bounded reads, and verified Drive, Docs, Sheets, Gmail, Calendar, or Chat writes.
handover
Write a self-contained follow-up prompt carrying this session's decisions, constraints, and verified findings into a fresh agent session. Use when handing unfinished work to a new context window.
dotfiles-bootstrap
Bootstrap a workstation with the dotfiles framework. Takes a GitHub user / owner+repo / explicit clone URL and runs dot init (which shells out to chezmoi) with the right safety prompts. Honors the active agent profile (ask / plan / apply / audit) so it defaults to dry-run in safer modes and full apply in apply.
vibe
Delegate a coding task to a cheap AI model (Mistral Vibe by default, but any provider Vibe knows about — DeepSeek, Gemini Flash, etc.) and supervise the result via git diff. Claude orchestrates, the cheap model codes. Claude consumes 500-1500 tokens per delegation regardless of how many file reads the delegate does…
telegram
Send notifications, interactive questions, or multiple-choice polls to the user via Telegram. Use when the user asks to be notified ("ping me", "notify me on Telegram", "ask me when..."), when a long-running task finishes and the user is likely away, when an irreversible action needs out-of-band confirmation, or when…
dotfiles
Chezmoi dotfiles conventions and patterns. Use when modifying dotfiles, adding new managed files, working with chezmoi templates, or syncing agent skills and subagents through agentspec. Do NOT use for installing tools, running doctor, or updating a machine through the dotfiles CLI; use manage-machine.