gtd-capture

gtd-capture is a skill for Claude Code from mikonos/LLM-GTD. It costs 54 tokens per session (1,320 once invoked), scanned A, original, MIT.

A GTD capture tool that records tasks, ideas, reminders, commitments, and unresolved questions in an inbox. For small batches, it can also sort them into the appropriate lists and define next steps.

In plain words
What is it for?
Use it to capture a single reminder, perform a brain dump, process the end of a work session, or automatically clarify a small number of new items.
Why use it?
It reduces the chance of losing thoughts while avoiding the need to organize everything before recording it. It can also turn a session’s commitments, blockers, and waiting items into organized GTD inputs.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the llm-gtd plugin — 8 skills, 8 commands shipped together

Good fit Use it to capture a single reminder, perform a brain dump, process the end of a work session, or automatically clarify a small number of new items.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/mikonos/llm-gtd/capture
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.

Any agent
npx skills add mikonos/LLM-GTD --skill capture
Clone the repo
git clone --depth 1 https://github.com/mikonos/LLM-GTD

Made for: Claude Code.

Or install llm-gtd, the plugin that ships this one along with the rest of its 8 skills, 8 commands.

Wrote 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.

agentmods badge for gtd-capture

README.md
[![agentmods](https://agentmods.dev/badge/skills/mikonos/llm-gtd/capture/github.svg)](https://agentmods.dev/skills/mikonos/llm-gtd/capture)
Your own site
<a href="https://agentmods.dev/skills/mikonos/llm-gtd/capture"><img src="https://agentmods.dev/badge/skills/mikonos/llm-gtd/capture/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.

agentmods 80×15 button for gtd-capture

Your own site · 80×15
<a href="https://agentmods.dev/skills/mikonos/llm-gtd/capture"><img src="https://agentmods.dev/badge/skills/mikonos/llm-gtd/capture.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 54 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,320 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00054 $0.01320
Opus 5 $0.00027 $0.00660
Sonnet 5 $0.00011 $0.00264
Haiku 4.5 $0.00005 $0.00132

Measured 8d ago against content hash 1e7beabd6ad4, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

gtd-capture 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.

plugins/llm-gtd/skills/gtd-harness/capture/SKILL.md · 54 lines

How it starts

The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.

GTD · capture(收集 → 默认自动理清)

视角:David Allen,AI-native 改良。Allen 当年把 capture / clarify 刻意分开,是为「人脑」——人切到决策模式有成本,所以捕捉只管零摩擦倒出来,理清留到专门 session。但 AI 切换成本≈0、能瞬间起草下一步,所以这里默认 capture 后自动 clarify——前提是不破坏 Allen 的两条真智慧:① 落盘永远第一步(先保证不丢);② 批量/mind-sweep 不逐条打断(高量时摩擦最致命)。

加载与边界

  • session 收尾:先读 templates/session-close-template.md
  • 单条默认自动 clarify:按 clarify/SKILL.md 执行;清单边界不明时读 references/list-definitions.md
  • 自动化边界:capture 可自动落 inbox;是否承诺、期望成果不清、行动/知识不清时问一句。

何时跑

  • 用户冒出待办、想法、提醒、承诺、未决问题、「别忘了…」。
  • 用户说「记一下」「丢进收件箱」「帮我捕捉」。
  • 用户要整理“今天 / 本轮 session 状态 / 本轮 session 收尾”:把会话里的承诺、阻塞、等待、硬日期、项目状态变化整理成 GTD 输入,并按固定五段输出。
  • 一次性倒空大脑(mind sweep):「我脑子太乱,帮我清空」。

工作流

  1. 零摩擦落盘(永远先做,保证不丢):把每项原样追加到 memory/gtd/inbox.md「待理清」段末尾,一项一行,保留原始措辞:

    - 原始措辞 · 捕捉日期:YYYYMMDD
    
  2. session 状态预处理(不是写日志):用户说“整理今天 / 本轮 session 状态 / 收尾”时,先读 templates/session-close-template.md,再把会话压缩成 3-5 条 GTD-relevant deltas:新承诺、待确认、等待别人、硬日期、项目状态变化或阻塞。每条按原始事实写成独立 inbox item;不要把完整聊天摘要、情绪流水账或知识洞察塞进 memory/gtd/。知识/洞察移交 ZK 管线;只有和活跃项目直接相关的支持材料才进 reference.md。收尾前跑 scripts/gtd_status.sh,最终按模板五段输出。

  3. 分流判断量级

    • 单条 / 少量(≤3 条)→ 默认自动 clarify:落盘后立即对每条跑 clarify 决策树(见 clarify/SKILL.md),把它理清归位到正确清单 + 定下一步,然后从 inbox 删除,一行回报:「已理清:〈原话〉→ 〈清单〉〈情境〉(下一步:〈具体动作〉)。要改归类/动作说一声。」——act-then-surface,不逐条问答(markdown 可秒改,乐观归位 + 易纠错优于事前盘问)。
    • 批量 / mind sweep(多条)→ 先全落盘,不逐条打断:用触发清单帮用户扫完、全丢 inbox,然后一次性问「要我现在逐条理清吗?」批量 clarify。保留 Allen 的零摩擦捕捉。
  4. 人类判断守卫(仅这几种才停下问一句,否则别打断)

    • 分不清「行动 vs 知识/想法」(知识该走 ZK 管线 fleeting-note,不进 GTD);
    • 明确是产品/功能/场景机会 → 自动归 product-ideas.md 保留原始机会,并默认同步生成 projects.md + next-actions.md 可见项,不混进 someday-maybe.md;只有用户明确说「先存不处理 / 只捕捉」才暂不升级;
    • 推不出具体下一步动作(信息不足);
    • 它隐含一个「你未必想承诺」的承诺(该不该做本身要你定);
    • 是 >1 步项目但期望成果不清。 这几种:先落盘,再问一句确认,不自行替用户拍板
  5. off-switch:用户说「只捕捉 / 先别理 / just record」→ 纯落盘 inbox,不自动 clarify,留给之后 /gtd-clarify

质量检查

  • 落盘永远先于理清(任何情况不丢)
  • 单条默认自动理清并归位 + 一行回报去向,可一句话纠错
  • mind-sweep 未被逐条打断(先全捕捉,再批量理清)
  • 只在四类「需你拍板」时停下问一句,其余不打断
  • session 状态只提取会影响承诺系统的变化,未制造第二套日志/日报,并按五段模板输出
  • 知识/想法类移交 ZK,未污染 GTD 清单

Read the full file on GitHub · 54 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. 8d ago First seen · 54 lines · 54 tokens per session scan A 1e7beabd6ad4

Subscribe to this mod's changes

gtd-capture is a skill published in the GitHub repository mikonos/LLM-GTD (10 stars, last pushed 2mo ago), licensed MIT. It adds 54 tokens to every session and 1,320 once invoked, about $0.0003 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.