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 PsychQuant/che-apple-mail-mcp --skill confirmation-protocolgit clone --depth 1 https://github.com/PsychQuant/che-apple-mail-mcpWrote 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/psychquant/che-apple-mail-mcp/confirmation-protocol)<a href="https://agentmods.dev/skills/psychquant/che-apple-mail-mcp/confirmation-protocol"><img src="https://agentmods.dev/badge/skills/psychquant/che-apple-mail-mcp/confirmation-protocol.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Agent Snooping · line 158 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 159 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- low Excessive Agency · line 145 Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.Fix: Limit the skill's scope to its documented purpose. Remove instructions that enable the agent to perform actions outside its stated functionality.
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 | $0.00088 | $0.01880 |
| Opus 5 | $0.00044 | $0.00940 |
| Sonnet 5 | $0.00018 | $0.00376 |
| Haiku 4.5 | $0.00009 | $0.00188 |
Grade A, and why
confirmation-protocol 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 8d 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 — 161 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Confirmation Protocol — Email Operations
核心原則
AI 不直接執行,先 show 結構化的「我理解你要的是這樣」讓 user confirm/correct/reject,achieve consensus 後才執行。
借鑑自 Foresay 的 Confirmation Protocol（該專案 2026-07 前名為 NSQL）。它最初設計給 SQL/data query;這裡 adapt 到 Apple Mail 操作。
Bootstrap(v2.9.0+ 鐵律,強制)
被 trigger 時(filter 模糊 / bulk operation ≥ 5 / destructive / compose 寄出),第一個動作就用 TaskCreate 把要跑的 phases 建成 todo list:
TaskCreate(subject="confirm_phase1_disambiguation",
description="Filter 是否模糊?模糊 → 列候選 + AskUserQuestion 選定;明確 → 直接 mark completed。處理完 update。")
TaskCreate(subject="confirm_phase2_search_preview",
description="Search 結果 ≥ 5 → bulk-operation-preview skill 跑 thread breakdown + false-positive flagging,展示 user;< 5 且無 ⚠⚠ → mark completed。")
TaskCreate(subject="confirm_phase3_operation_confirmation",
description="展示 side-effect scope (N files written, K attachments, total bytes)。Destructive 加 ⚠ 警告。User AskUserQuestion 答覆(a)確認 (b)排除 false positive (c)改 filter (d)取消。")
TaskCreate(subject="confirm_phase4_execute_or_iterate",
description="按 user 答覆執行 / 回 Phase 1-3 / abort。Abort 時要明確告訴 user『未做任何 side effect』。")
完成每一個 phase 立即 TaskUpdate → completed。靜默 skip = 違規。即使該 phase 是「可 skip」(明確 filter 跳 Phase 1),也要 TaskUpdate completed + 註明 skip 原因(在 description 裡 append),不可只是不做。
為什麼:沒 task tracking 時,Claude 容易把 Phase 2 preview 用「應該還好吧」rationalize 掉,直接寫檔。Task 標 incomplete 在 UI 看得見,user 知道有東西被跳。
何時觸發
| 觸發條件 | 範例 | 嚴重度 |
|---|---|---|
| Filter ambiguity | 「陳老師」、「最近的信」、「VIP 寄來的」 | 🔴 必 confirm |
| Bulk operation | search 結果 ≥ 5 封 | 🟡 建議 confirm |
| Destructive | delete_email、empty Trash、bulk move | 🔴 必 confirm |
| Compose 寄出 | compose_email、reply_email | 🔴 必 confirm |
| Outbound side effect | mark_as_junk、forward_email | 🟡 建議 confirm |
Skip confirm(可以直接執行):
- 明確 Message-ID 的 read-only 操作(get_email、list_attachments)
- Single email mark_read/unflag
- 純 search/list(沒有 side effect)
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.
- 8d ago First seen · 161 lines · 88 tokens per session scan A a9f5fa7d5bf5
confirmation-protocol is a skill published in the GitHub repository PsychQuant/che-apple-mail-mcp (8 stars, last pushed today), licensed MIT. It adds 88 tokens to every session and 1,880 once invoked, about $0.0004 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.
Other skills, from other repositories
apple-mail
Use this skill when the user wants to manage Apple Mail on macOS - reading, searching, sending, replying to, forwarding, and organizing emails and mailboxes. This skill provides access to Apple Mail through MCP tools.
email-management-expert
Expert email management assistant for Apple Mail. Use this when the user mentions inbox management, email organization, email triage, inbox zero, organizing emails, managing mail folders, email productivity, checking emails, or email workflow optimization. Provides intelligent workflows and best practices for…
apple-notes
Use this skill when the user wants to interact with Apple Notes on macOS - creating, searching, reading, updating, deleting, organizing, or formatting notes and folders. This skill provides access to Apple Notes through MCP tools and includes safe formatting guidance.
archive-event
Archive a meeting or event into Calendar from a narrative source — a meeting notice, an announcement, a message thread — deriving the time from the thread's authoritative correction, recording the source in the event notes, and labelling any value the source did not state. Use when the user hands over a notice…
calendar-management
Guide for effective use of che-ical-mcp for macOS Calendar & Reminders management. Use when user asks about calendar events, reminders, scheduling, or time management.
c-email
Read, send, search, and label email via gog (Gmail CLI) or himalaya (IMAP). Supports inbox management, drafting replies, thread viewing, and label/folder operations across Gmail and standard IMAP accounts.