菜单
Read and write durable agent memory in the personal Obsidian vault at $AGENT_MEMORY_VAULT. Use at session start to recall context, when the user states a lasting preference, decision, or fact, or when asked to remember, recall, or forget something.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Design, create, and validate portable Agent Skills with effective triggers, concise instructions, and reusable scripts, references, or assets. Use when adding a new skill, substantially restructuring an existing skill, or validating a skill's structure, frontmatter, links, and bundled scripts.
Manage GitHub via CLI including pull requests, issues, workflows, actions, releases, and repositories. Use when working with GitHub, gh commands, GitHub Actions CI/CD, PRs, issues, releases, or repository management.
Check reasoning with a second mind — dispatch a reviewer whose prompt is built for the specific question, with a lens it can fail on and evidence required for every finding. Use before acting on a plan, decision, diagnosis, or rationale whose cost of being wrong is high and whose only support so far is your own reasoning.
Use tmux safely from an agent with explicit pane targeting, state inspection, verified input delivery, response polling, recovery, and multi-agent supervision. Use when an agent CLI, TUI, REPL, authentication flow, or other interactive command must retain terminal state across tool calls.
Confirm you have everything needed to finish a change before starting it — what to implement, how it will be tested, and which rules govern it. Use before non-trivial work, when reviewing a plan for completeness, or when preparing a PR that needs planning evidence.
Manage Linear issues, teams, and projects via the Linear CLI. Use when the tracker is Linear specifically — reading or filing Linear issues, Linear sprint workflows, creating a branch or PR from a Linear issue, or managing Linear teams and projects. Not for GitHub Issues or other trackers.
| name | memory-conventions |
| description | Read and write durable agent memory in the personal Obsidian vault at $AGENT_MEMORY_VAULT. Use at session start to recall context, when the user states a lasting preference, decision, or fact, or when asked to remember, recall, or forget something. |
Durable, cross-harness agent memory lives in the vault at
$AGENT_MEMORY_VAULT. Access is direct file operations — no CLI or
app required, works headless on every harness.
The agent/ directory is an OKF-conformant bundle (markdown + YAML
frontmatter; permissive consumption — never reject a malformed fact).
$AGENT_MEMORY_VAULT/
agent/
index.md # the only file loaded at session start
log.md # append-only history (temporal layer)
facts/<slug>.md # one fact per note, semantic kebab-slug filename
Read agent/index.md only (hard cap 200 lines / 25KB). Load individual
facts on demand when their hook matches the task.
index.md is a pure listing (OKF §6): a heading and link entries,
nothing else. Its only frontmatter is okf_version: "0.1" at the
bundle root. Conventions prose lives here in this skill, never in the
index.
When the user states something durable (preference, decision, constraint, reference):
agent/facts/<kebab-slug>.md — the slug names the
concept, never a timestamp (concept = identity; update in place):---
type: user|feedback|project|reference
title: <display name>
description: <one line — becomes the index hook>
created: <ISO 8601 with seconds, UTC, e.g. 2026-07-12T18:42:07Z>
updated: <same format>
source: <where this came from>
---
<the fact, one per note; absolute dates; [[wiki-links]] to related facts>
facts/<slug>.md, followed by
— <description>.agent/log.md under a ## YYYY-MM-DD heading (newest
day first). Entry: bold operation verb (**Create**, **Update**,
**Delete**), the fact's title linked to its file, — <reason>,
and the UTC time (HH:MM:SSZ).sync doctor enforces this).$AGENT_MEMORY_VAULT is unset or missing, say so and skip memory
operations; do not invent a location.