URL Manager for Openclaw

An agent-first, cross-platform bookmarking and knowledge management tool that transforms raw links into structured and searchable card-based collections.

piccolo123
v1.0.0
Jul 21, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install sync-test

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install sync-test using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is URL Manager?

URL Manager is a developer-focused knowledge management tool designed to transition web links from raw text into structured, card-based collections. Tailored for agents executing tasks within the Openclaw Skills framework, this utility establishes a clean delivery layer where saved resources are organized, tagged, and ready for human consumption. This replaces traditional chat-based URL dumps with a dedicated user interface.

By serving as an interface between AI agents and users, the tool allows agents to register accounts, generate secure magic links, and update collections. Operating over Python and integrating through MCP servers, this tool offers a structured approach to how users and AI agents co-curate digital resources.

URL Manager Use Cases

  • Structured Bookmarking: Storing and cataloging web links under specific category sets and tags instead of posting raw URLs in chat.
  • Automated Research Curation: Enabling AI agents utilizing Openclaw Skills to group collected research documents, code repositories, or reference articles on the fly.
  • Team Knowledge Base Shared Cocreation: Setting up multi-editor shared categories where multiple team members or distinct agents collaborate on curated sets.
  • Resource Delivery: Providing users with a clean, card-based overview via secure, single-use magic links instead of linear chat logs.

How URL Manager Works

  1. Agent Registration: The AI agent initiates account creation using local scripts or MCP protocols, generating an authorization token.
  2. Resource Parsing: The agent extracts metadata such as the URL, title, description, and tags from the reference material.
  3. Database Insertion: The resource is stored as a footprint using API endpoints, mapped to specified category sets.
  4. Structured Delivery: The agent requests a single-use magic link from the platform and presents it to the user.
  5. Client Rendering: The user opens the link to view a card-based interface containing all curated links, organized as structured by the agent.

URL Manager Setup

Path A: MCP Protocol Connection

To connect the tool to an MCP-compatible environment such as Cursor, Claude Desktop, or Openclaw Skills, use the following server configuration:

{
  "mcpServers": {
    "url-manager": {
      "command": "python",
      "args": ["server.py"],
      "env": {
        "FOOTPRINTS_TOKEN": ""
      }
    }
  }
}

When FOOTPRINTS_TOKEN is blank, call agent_register() to generate and activate the credentials.

Path B: Local CLI Execution

Configure environment variables for non-MCP runtimes:

export FOOTPRINTS_TOKEN="FA_xxxxxxxxxxxx"
export FOOTPRINTS_ENDPOINT="https://ai.ocean94.com"

Run the agent registration tool if the user does not possess an existing account:

python3 scripts/footprints.py agent_register

To insert a new resource footprint via CLI:

python3 scripts/footprints.py footprints_add \
  --url "https://example.com/article" \
  --title "Article Title" \
  --description "Extracted summary of target URL" \
  --category-ids 1,2 \
  --tags "AI,tutorial"

URL Manager Data Schema & Taxonomy

URL Manager structures its metadata catalog via specific taxonomies to keep bookmarks organized:

Concept Description Cardinality / Formatting
Footprint A registered link containing title, description, tags, and category associations. Core data entity
Category Logical grouping bucket for footprints. Many-to-many relationship
Category Set Top-level workspace container (e.g., 'Work', 'Personal'). Contains multiple Categories
Tag Lightweight keyword label to facilitate quick filtering. Comma-separated strings
Shared Category A cooperative collection that supports reader-only subscription or multi-editor cocreation. Shared via invite link

Security Note: All read and write operations are audited. Logs capture the timestamp, event type, and target footprint, ensuring transparent agent behavior.

URL Manager Advanced Features

  • Agent-First Auto-Registration: Background account creation that allows tools integrated with Openclaw Skills to register, operate, and deliver content with minimal user overhead.
  • Single-Use Magic Links: Delivery of single-use, 15-minute secure magic links, presenting an alternative to raw link dumps inside chat streams.
  • Multi-Agent Collaborative Collections: Shared categories that enable distinct agents to co-curate single or multiple reference datasets.
  • Batch Updates and Rate Mitigation: Native bulk-updating workflows designed to manage API rate limits when updating numerous links simultaneously.
  • Automatic User Account Merging: Conflict resolution that transitions agent-created accounts to registered user profiles seamlessly.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*