← Back to Skills Marketplace
1241
Downloads
0
Stars
17
Active Installs
1
Versions
Install in OpenClaw
/install fetch
Description
Public web retrieval and clean extraction engine. Use whenever the user wants to fetch, download, inspect, clean, or save content from a public URL. Supports...
README (SKILL.md)
Fetch
Turn public URLs into usable local content.
Core Philosophy
- Fetch only public web content.
- Prefer clean extracted text over noisy raw HTML.
- Save both the raw response and structured extraction locally.
- Keep a simple local job history so previous fetches are easy to inspect.
Runtime Requirements
- Python 3 must be available as
python3 - No external packages required
Safety Boundaries
- Public URLs only
- No login flows
- No cookies or browser automation
- No API keys or credentials
- No external uploads or cloud sync
- All fetched data is stored locally only
Local Storage
All data is stored under:
~/.openclaw/workspace/memory/fetch/jobs.json~/.openclaw/workspace/memory/fetch/pages/
Key Workflows
- Fetch URL:
fetch_url.py --url "https://example.com" - Save cleaned output:
save_output.py --url "https://example.com" --title "Example" - List history:
list_jobs.py - Show job details:
show_job.py --id JOB-XXXX
Scripts
| Script | Purpose |
|---|---|
init_storage.py |
Initialize local storage |
fetch_url.py |
Fetch a public URL and extract content |
save_output.py |
Save cleaned output with a custom title |
list_jobs.py |
List previous fetch jobs |
show_job.py |
Show one saved fetch job |
Usage Guidance
This skill appears coherent and self-contained. Before installing: (1) review and, if desired, run the included scripts locally to confirm behavior; (2) be cautious when fetching untrusted URLs — large responses are not size-limited and raw HTML may contain sensitive data you wouldn't want stored locally; (3) confirm you are comfortable with files being created under ~/.openclaw/workspace/memory/fetch. No credentials or external uploads are requested by the skill.
Capability Analysis
Type: OpenClaw Skill
Name: fetch
Version: 1.0.0
The skill bundle provides a legitimate utility for fetching and cleaning public web content using standard Python libraries (urllib, html.parser). It implements basic safety checks, such as URL scheme validation in fetch_url.py, and stores all data locally within a designated workspace directory (~/.openclaw/workspace/memory/fetch/). No evidence of data exfiltration, shell injection, or malicious prompt instructions was found.
Capability Assessment
Purpose & Capability
Name/description (public fetch + clean + local save) align with the provided scripts: fetch_url.py performs an HTTP(S) GET, extract.py cleans/extracts title/links, and storage writes files under ~/.openclaw/workspace/memory/fetch. There are no requests for unrelated credentials or services.
Instruction Scope
SKILL.md instructions match the scripts' behavior: they require python3, operate on public URLs, store data locally, and offer list/show/save workflows. The scripts do not read other system files, contact endpoints beyond the target URL, or perform browser automation. Minor note: extracted links are returned as-is (may include non-http schemes) and large downloads are not size-limited.
Install Mechanism
No install spec and no external package downloads — the skill is delivered as scripts and uses only Python stdlib. This is the lowest-risk install model.
Credentials
The skill requires no environment variables, credentials, or config paths beyond writing to its own ~/.openclaw workspace. Declared runtime constraints (no cookies, no logins) match the code.
Persistence & Privilege
always is false; the skill does not request permanent/global agent privileges or modify other skills. It writes only to its own workspace directory and job file.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install fetch - After installation, invoke the skill by name or use
/fetch - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
[email protected]: Public web retrieval and clean extraction engine. Fetch public URLs, extract readable text, save raw and cleaned output locally, and keep a simple local job history.
Metadata
Frequently Asked Questions
What is Fetch?
Public web retrieval and clean extraction engine. Use whenever the user wants to fetch, download, inspect, clean, or save content from a public URL. Supports... It is an AI Agent Skill for Claude Code / OpenClaw, with 1241 downloads so far.
How do I install Fetch?
Run "/install fetch" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Fetch free?
Yes, Fetch is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Fetch support?
Fetch is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Fetch?
It is built and maintained by AGIstack (@agistack); the current version is v1.0.0.
More Skills