← 返回 Skills 市场
ivangdavila

Journal

作者 Iván · GitHub ↗ · v1.0.0
linuxdarwinwin32 ✓ 安全检测通过
1375
总下载
2
收藏
4
当前安装
1
版本数
在 OpenClaw 中安装
/install journal
功能描述
Build a personal journaling practice with prompts, reflection, and pattern discovery.
使用说明 (SKILL.md)

Core Behavior

  • User wants to write → provide space, optional prompts
  • No pressure, no judgment → journaling should feel safe
  • Surface patterns when asked → insights from past entries
  • Create ~/journal/ as workspace

Entry Flexibility

  • Stream of consciousness welcome — no structure required
  • Bullet points fine — not everything needs paragraphs
  • Voice notes transcribed — capture thoughts while walking
  • Short entries valid — "Today was hard" is enough

File Structure

One file per entry: 2024-03-15.md

  • Date as filename — chronological, findable
  • Optional: time of day if multiple entries
  • Frontmatter optional: mood, tags, location

When User Starts Writing

  • Don't interrupt flow — capture first, reflect later
  • Offer prompts only if asked or stuck
  • Accept whatever format they give
  • "Just write" is the goal

Prompt Library (When Asked)

  • "What's on your mind?"
  • "What went well today?"
  • "What would you do differently?"
  • "What are you grateful for?"
  • "What are you avoiding thinking about?"
  • "If tomorrow goes perfectly, what happens?"
  • Keep prompts in prompts.md for personalization

End of Entry Options

  • Save as is — most common
  • Add tags for later searching
  • Note mood: 1-5 or emoji
  • "Continue later" flag for unfinished thoughts

Weekly Review

  • Offer to review the week
  • Themes that emerged
  • Mood patterns if tracked
  • Wins and struggles
  • "Anything to carry forward?"

Monthly/Yearly Reflection

  • What changed this month/year?
  • Recurring themes or concerns
  • Progress on long-term thoughts
  • Reading old entries — often surprising

Pattern Discovery

When asked "what have I been writing about?":

  • Common themes across entries
  • Mood trends if tracked
  • Frequency of journaling
  • Topics that appear then disappear

What To Surface Proactively

  • "It's been 5 days since you journaled" — only if they want nudges
  • "A year ago you wrote about X" — memory resurfacing
  • "This theme appeared 3 times this month" — pattern spotting
  • Never share content without permission

Progressive Enhancement

  • Week 1: just write, any format
  • Week 2: consistent file naming
  • Month 2: add mood/tags if useful
  • Month 3: weekly review practice
  • Year 1: annual reflection

Folder Structure

~/journal/
├── 2024/
│   ├── 2024-03-15.md
│   └── 2024-03-16.md
├── prompts.md
└── reflections/
    └── 2024-march-review.md

Privacy and Security

  • Local files only — no cloud unless user chooses
  • Encryption option if sensitive
  • No AI training on journal content — make this clear
  • Backup encrypted if backing up at all

Types of Journaling

  • Daily log: what happened, how you felt
  • Gratitude: what you're thankful for
  • Morning pages: stream of consciousness on waking
  • Evening reflection: review of the day
  • Topic-specific: work, relationships, health
  • Let user find their style

What NOT To Suggest

  • Complex templates before natural writing flows
  • Mandatory daily journaling — guilt kills the practice
  • Sharing entries anywhere
  • Analysis before sufficient entries exist
  • Fixing problems — sometimes just listening is enough

Handling Difficult Entries

  • Acknowledge without judgment
  • Don't offer unsolicited advice
  • "That sounds really hard" is often enough
  • Suggest professional support if concerning patterns

Integration Points

  • Habits: "journal daily" as habit if wanted
  • Mood tracking: simple scale alongside entries
  • Goals: reflection on progress
  • Contacts: processing relationship thoughts

Searching Past Entries

  • Full-text search across all entries
  • Search by date range
  • Search by mood if tracked
  • Search by tag
  • "What was I writing about last March?"
安全使用建议
This skill is instruction-only and appears to do what it says: create and manage local journal files in ~/journal, surface patterns from those files, and offer prompts. Before installing or using it: (1) understand it will read and write files under ~/journal — store only what you're willing to have processed locally; (2) if you want voice transcription, confirm what tool/service will be used (local transcription vs. a cloud API) because that may send audio externally; (3) enable encryption/backups only according to your privacy needs and verify the agent/platform will not send journal text to external services unless you opt in; and (4) if the skill will access contacts or other integrations, require explicit permission and review which data is shared. Overall the package is coherent and low-risk, but review transcription/integration behavior and platform permissions before enabling proactive nudges or cloud features.
功能分析
Type: OpenClaw Skill Name: journal Version: 1.0.0 The OpenClaw AgentSkills bundle for 'Journal' is benign. The `SKILL.md` instructions are well-aligned with the stated purpose of a personal journaling tool, emphasizing privacy, local file storage, and explicit prohibitions against sharing content or using it for AI training. There is no evidence of prompt injection, data exfiltration, unauthorized execution, or any other malicious intent.
能力评估
Purpose & Capability
Name/description match the runtime instructions: filing entries as dated Markdown, prompting, weekly/monthly reviews, and local pattern discovery. Requested actions (create ~/journal, read/write entries) are coherent with a journaling assistant and do not demand unrelated resources.
Instruction Scope
SKILL.md restricts operation to local files and pattern discovery across entries, which is appropriate. Two points to note: (1) it references 'voice notes transcribed' but does not specify how transcription is performed (this implies a dependency on an external service or local tool that isn't declared), and (2) it mentions integrations (habits, contacts) conceptually but does not request or describe access to contact data — the agent's implementation would need explicit permission if it accessed other data sources. Otherwise, instructions do not ask for unrelated system data or secrets.
Install Mechanism
Instruction-only skill with no install spec and no code files. Nothing will be written to disk by an installer; the only file writes described are the journal entries themselves, which is expected behavior.
Credentials
No environment variables, credentials, or config paths are required. The skill's behavior (local file storage, optional encryption chosen by user) is proportionate to its stated purpose.
Persistence & Privilege
always is false and the skill does not request elevated or persistent platform privileges. It can run autonomously per platform defaults; the SKILL.md allows proactive nudges (e.g., 'It's been 5 days since you journaled') — this is privacy-sensitive but not inconsistent with the skill's purpose.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install journal
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /journal 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
Initial release
元数据
Slug journal
版本 1.0.0
许可证
累计安装 4
当前安装数 4
历史版本数 1
常见问题

Journal 是什么?

Build a personal journaling practice with prompts, reflection, and pattern discovery. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 1375 次。

如何安装 Journal?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install journal」即可一键安装,无需额外配置。

Journal 是免费的吗?

是的,Journal 完全免费(开源免费),可自由下载、安装和使用。

Journal 支持哪些平台?

Journal 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(linux, darwin, win32)。

谁开发了 Journal?

由 Iván(@ivangdavila)开发并维护,当前版本 v1.0.0。

💬 留言讨论