memory-deliver

memory-deliver is a skill for Claude Code, Codex from Meari-Prototype/iftree-editor. It costs 91 tokens per session (1,364 once invoked), scanned A, original, Apache-2.0.

A session-memory tool that saves the conversation record to an external memory library and drafts selected long-term notes for human review.

In plain words
What is it for?
Use it at the end of a session, or when a user says to remember something, to save the source conversation and prepare a note for approval.
Why use it?
It keeps important context from being lost between sessions while avoiding unreviewed changes to long-term memory.

Skill for Claude CodeCodex

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.

agentmods
npx agentmods add skills/meari-prototype/iftree-editor/memory-deliver
Any agent
npx skills add Meari-Prototype/iftree-editor --skill memory-deliver
Clone the repo
git clone --depth 1 https://github.com/Meari-Prototype/iftree-editor

Made for: Claude Code, Codex.

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 memory-deliver

README.md
[![agentmods](https://agentmods.dev/badge/skills/meari-prototype/iftree-editor/memory-deliver.svg)](https://agentmods.dev/skills/meari-prototype/iftree-editor/memory-deliver)
Your own site
<a href="https://agentmods.dev/skills/meari-prototype/iftree-editor/memory-deliver"><img src="https://agentmods.dev/badge/skills/meari-prototype/iftree-editor/memory-deliver.svg" alt="Measured on agentmods" height="20"></a>
Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,364 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00091 $0.01364
Opus 5 $0.00046 $0.00682
Sonnet 5 $0.00018 $0.00273
Haiku 4.5 $0.00009 $0.00136

Measured 4d ago against content hash 4e76bd0c3c04, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

memory-deliver 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 4d 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.

.iftree-llm-workspace/skills/memory-deliver/SKILL.md · 69 lines

How it starts

The opening of the file, as written. The whole thing — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.

事件卷投递 + 长期记忆起草(projectneed 18-8-4 / 15-10 / 15-11

把你这个会话投进本产品的记忆。记忆系统除路径外没有任何特殊机制——事件卷、长期记忆都是普通条件树文档,全走主系统已有的导入 / 流式写入 / 编辑分支 / 引用 / 合并,只多几条规则让它像记忆。本文写给执行投递的外部 LLM(Claude Code、Codex、任何能调 MCP 的 agent)。

两件事

  1. 投递事件卷(每会话):把这个 session 的宿主原始记录导入成卷——"确曾发生过什么"。你不复述、不整理:系统读 hostAnchor 指向的真实 session 文件、纯规则解析成卷(确定可重复)。你只给三个东西:agentsessionIdhostAnchor
  2. 起草长期记忆(值得时):把"现在如此"的可复用结论,在长期记忆文档上开 draft 提 diff、连到事件卷里那句出处原话,人审落地。一个 session 常零产出、顶多一两条。

原则

  • 事件卷是原料、不是结论。 卷答"发生过什么";一律 trust 不受控(15-10-3)。你不借它固化判断。
  • 一 session 一卷。 同 agent+sessionId 重投 = 旧卷全删 + 完整重导(session 只追加 + 解析确定 → 重导 ≡ 追加),不产生第二个卷、不留孤儿。你重复投递无害。
  • 长期记忆经人审、不直写。 你在记忆文档上开 draft 提 diff(15-11),人审落地——这就是"起草待审",不是直写结论。每条回指事件卷出处(用户真说的那句话)。
  • 除路径外无特殊机制。 全走主系统现成动词(import / push / draft / edit / ref / commit / merge);不必学新东西。

投递时机

  • 会话收尾(主):结束前投一次事件卷。
  • 用户说"记一下"/"你记住了":当场投事件卷 + 起草那条长期记忆。

1) 投递事件卷

memory_deliver { agent, sessionId, hostAnchor, title?, startedAt?, endedAt? }
  • agent:稳定的框架标识(claude-code / codex),同框架别换写。
  • sessionId:宿主侧 session id。
  • hostAnchor:你这个 session 的真实 transcript 文件路径(路径#sessionid)。系统读它解析成卷;文件不存在 / 解析不出对话即拒(不空卷直接造)。
  • 不传 nodes:系统规则解析你的 session 文件,你不整理自述日志。

命令行:db memory deliver <json>(同契约)。投完用 memory_volumes / db memory list 确认卷已落库。

2) 起草长期记忆(值得时)

值得长期记住的(用户明确"记一下"、踩过的坑、可复用结论),在长期记忆文档上起草——不是直写,是开草稿待人审:

  1. 找到你身份的长期记忆文档(library/memory/<身份>.md,每身份一篇;没有就先建一篇)。
  2. draft 起草 → edit 写一条精炼的记忆节点("现在如此")→ editref.addNodeToNode 把它连到刚投事件卷里那句用户原话节点(出处)。
  3. commit 提交待人审(human 档 merge 落主干)。

每条记忆必须是用户真说过 / 真踩过、且能回指事件卷出处。一 session 常零产出、顶多一两条——一条都算多,宁缺毋滥。

自检

  • 事件卷只给 agent / sessionId / hostAnchor,hostAnchor 指向真实 session 文件
  • 长期记忆是起草待审(draft + 人审)、非直写;每条回指事件卷出处
  • 长期记忆条目是用户真说 / 真踩过的,精炼成"现在如此"

常驻指令模板(贴进 CLAUDE.md / AGENTS.md / NOW.md)

底线档靠它生效——没有这段,你不会记得投递:

## 记忆(IFTree 库)
- 跨会话记忆:开工先看最近发生过什么(memory_volumes 列卷,find/tree/read 下钻),
  再看长期核心记忆与工作区文件;采信前看时间元数据,同主题冲突新者胜。
- 会话收尾投事件卷(memory_deliver,只给 agent/sessionId/hostAnchor,系统读 session 文件导入,
  契约见 .iftree-llm-workspace/skills/memory-deliver/SKILL.md);
  用户说"记一下"时当场投 + 起草那条长期记忆(draft 待人审)。
- 长期记忆经起草待审、不直写结论;每条回指事件卷出处。

Read the full file on GitHub · 69 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. 4d ago First seen · 69 lines · 91 tokens per session scan A 4e76bd0c3c04

Subscribe to this mod's changes

memory-deliver is a skill published in the GitHub repository Meari-Prototype/iftree-editor (5 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 91 tokens to every session and 1,364 once invoked, about $0.0005 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.

Related

Other skills, from other repositories

karpathy-llm-wiki

Use when building or maintaining a personal LLM-powered knowledge base. Triggers: ingesting sources into a wiki, querying wiki knowledge, linting wiki quality, 'add to wiki', 'what do I know about', or any mention of 'LLM wiki' or 'Karpathy wiki'.

Astro-Han/karpathy-llm-wiki · 67 tokens

openkb

Use when the user asks about content in their OpenKB knowledge base — research topics, concepts compiled from their documents, cross-document synthesis — or mentions openkb, an .openkb/ directory, or a wiki/ tree generated by openkb. The user may invoke you from any working directory; the active KB resolves via openkb…

VectifyAI/OpenKB · 102 tokens

link-memory

Use after important user-approved decisions, when durable context should be proposed or reviewed, and for explicit Link memory lifecycle work: remember, recall, review, update, archive, restore, forget, or explain local memories through the CLI without requiring MCP.

gowtham0992/link · 0 tokens

link-retrieve

Use before answering work that may depend on user memory, project history, source-backed notes, or prior decisions; retrieve compact Link context through the CLI without loading the whole wiki or requiring MCP.

gowtham0992/link · 42 tokens

link-ingest

Use when raw files are present, source pages look stale, or a user asks to ingest notes into Link; refresh source-backed wiki pages, propose memories, and validate updates through the CLI without MCP.

gowtham0992/link · 44 tokens

mindbase

Maintain a persistent AI-tended research wiki. Activates when the user wants to ingest a source (PDF, URL, article, paste), capture a thought, build the wiki context, query what's already known, lint for orphans/contradictions, or maintain page conventions. Phrases include "add this to my wiki", "today I decided…"…

frankchu91/mindbase-llm-wiki · 105 tokens