← 返回 Skills 市场
991
总下载
0
收藏
4
当前安装
1
版本数
在 OpenClaw 中安装
/install archive
功能描述
Capture and preserve content as intelligent snapshots with semantic search, automatic extraction, and proactive resurfacing.
使用说明 (SKILL.md)
Architecture
Archive storage lives in ~/archive/ with tiered structure. See memory-template.md for setup.
~/archive/
├── memory.md # HOT: recent items, ≤100 lines
├── index.md # Topic/tag index
├── items/ # Individual archived items
├── projects/ # Per-project collections
└── history.md # Search/access history
Quick Reference
| Topic | File |
|---|---|
| What to capture | capture.md |
| Search patterns | search.md |
| Resurfacing rules | resurface.md |
Core Rules
1. Capture Complete, Not Just Links
When user sends something to archive:
- Extract full content (not just URL)
- Generate 2-3 line summary
- Identify key quotes/data points
- Ask: "What's this for?" — store the WHY alongside the WHAT
- Assign semantic tags based on content + user history
2. Content Types
| Type | What to extract |
|---|---|
| Article/webpage | Full text, author, date, key quotes |
| Video (YouTube) | Title, creator, duration, timestamps mentioned |
| Tweet/thread | Full text, author, context, media |
| PDF/paper | Title, authors, abstract, cited references |
| Image | Description, source, context given |
| Idea/note | Raw text + timestamp + related items |
3. Storage Structure
Each archived item stored as:
items/{date}_{slug}.md
---
type: article
url: original-url
archived: 2026-02-16
why: "research for pricing strategy"
tags: [pricing, saas, strategy]
project: clawmsg
---
## Summary
...
## Key Points
...
## Full Content
...
4. Semantic Search
User can ask naturally:
- "What did I save about X?" → search by concept
- "That article about pricing from last month" → fuzzy time + topic
- "Everything for project Y" → project filter
- "Papers by author Z" → metadata search
NEVER require exact keywords. Match by meaning.
5. Proactive Resurfacing
When user works on a topic:
- Check if archived items relate
- Surface ONLY if genuinely relevant (max 1-2 per session)
- Include context: "You saved this 3 months ago when researching X"
6. Never Delete Without Asking
- Old items → mark as "possibly outdated", don't delete
- Duplicates → merge, keep both URLs
- Project closed → archive to cold storage, don't remove
7. Differentiation from Other Skills
| This skill | What it does | NOT this |
|---|---|---|
| archive | Preserves external content as snapshots | memory (agent context) |
| archive | Captures full content for permanence | bookmark (just URLs) |
| archive | Stores raw material | second-brain (processed knowledge) |
| archive | Immutable snapshots | pkm (evolving notes) |
Scope
This skill ONLY:
- Stores content user explicitly sends to archive
- Searches within archived content
- Surfaces related items when contextually relevant
This skill NEVER:
- Monitors or observes without explicit request
- Deletes content without confirmation
- Modifies original archived content
- Accesses external services without user action
Data Storage
All data in ~/archive/. Create on first use:
mkdir -p ~/archive/items ~/archive/projects
安全使用建议
This skill appears to be what it says: an on-device archiver that stores items under ~/archive/ and provides semantic search/resurfacing. Before installing, consider: 1) Network fetching: when you ask it to archive a URL, the agent will fetch remote content — confirm whether that fetch happens locally or via any third-party tool (to avoid unintended content leakage). 2) Local storage: everything is stored in ~/archive/ (full text, metadata and any local file paths) — avoid archiving secrets, credentials, or private files you don't want persisted. 3) Semantic processing: the skill is vague about how embeddings/searching are implemented; if it uses an external service you may need to supply API keys (none are requested now), so verify where processing occurs. 4) Resurfacing: proactive suggestions are limited but enabled by default; use the opt-out commands described if you prefer search-only. If you want higher assurance, ask the publisher how web fetches and semantic indexing are performed and whether any third-party endpoints are contacted during capture or search.
功能分析
Type: OpenClaw Skill
Name: archive
Version: 1.0.0
The OpenClaw Archive skill is classified as benign. Its instructions and code are clearly aligned with its stated purpose of capturing and preserving content. The skill explicitly defines ethical boundaries, stating it 'NEVER Monitors or observes without explicit request,' 'NEVER Deletes content without confirmation,' and 'NEVER Accesses external services without user action' in `SKILL.md`. File system operations (`mkdir`, `touch` in `memory-template.md`) are limited to creating local directories (`~/archive/`) for data storage, which is a standard and expected behavior for such a skill. While fetching web content (`web_fetch or browser` in `capture.md`) involves network access, it is essential for the skill's core functionality and is constrained by the skill's own rules, indicating no malicious intent.
能力评估
Purpose & Capability
Name/description (archive + semantic search + resurfacing) match the instructions: create ~/archive/, extract and store content, provide semantic search and limited proactive suggestions. No unrelated credentials, binaries, or install steps are requested.
Instruction Scope
Instructions stay largely within the stated purpose (capture, store, search, resurface). They do instruct the agent to 'fetch full content (web_fetch or browser)' which implies network access when the user requests URL capture; the SKILL.md also references using 'user history' and semantic matching but does not instruct reading unrelated system files or secret env vars. Note: the mechanism for generating semantic matches (local model vs external embedding service) is unspecified.
Install Mechanism
No install spec or code files — instruction-only skill. Nothing will be downloaded or written by an installer beyond the archive directory the skill itself creates on first use.
Credentials
No required environment variables, credentials, or config paths are declared. The requested access (filesystem under ~/archive/ and optional local PDF paths) is proportional to the stated functionality.
Persistence & Privilege
Not always:true and does not request elevated or cross-skill configuration. It writes/reads only its own archive files per instructions and describes opt-out behavior for resurfacing.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install archive - 安装完成后，直接呼叫该 Skill 的名称或使用
/archive触发 - 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
常见问题
Archive 是什么？
Capture and preserve content as intelligent snapshots with semantic search, automatic extraction, and proactive resurfacing. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 991 次。
如何安装 Archive？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install archive」即可一键安装，无需额外配置。
Archive 是免费的吗？
是的，Archive 完全免费（开源免费），可自由下载、安装和使用。
Archive 支持哪些平台？
Archive 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 Archive？
由 Iván（@ivangdavila）开发并维护，当前版本 v1.0.0。
推荐 Skills