A professional framework for building local-first retrieval architectures with strict security boundaries and automated maintenance workflows.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install workspace-local-retrieval
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install workspace-local-retrieval 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 Workspace Local Retrieval?
Workspace Local Retrieval is a sophisticated architectural pattern designed for Openclaw Skills to implement Retrieval-Augmented Generation (RAG) without the chaos of unmanaged indexing. It solves common local search failures by enforcing explicit corpus boundaries and a deny-by-default access policy for AI agents. This ensures that sensitive data remains protected while providing agents with the specific context they need to perform tasks accurately.
By separating personal memory from reusable workspace knowledge, this skill provides a stable, maintainable foundation for multi-agent environments. It prioritizes local-first storage and incremental refresh cycles, making it an ideal choice for developers who value privacy, explainable search status, and professional-grade data governance within their AI workflows.
Workspace Local Retrieval Use Cases
- Implementing local RAG without indexing irrelevant or private files.
- Separating personal memory layers from shared workspace knowledge bases.
- Defining granular, agent-scoped access to specific document corpora.
- Transitioning one-off local search scripts into reusable, packageable AI skills.
- Managing large-scale document sets with maintenance-aware incremental refreshes.
- Enforcing preflight environmental checks to ensure search reliability across different operating systems.
How Workspace Local Retrieval Works
- Preflight Validation: The system runs a mandatory check of hardware and software dependencies to ensure the environment is ready for embedding and indexing.
- Boundary Definition: Users define explicit corpora allowlists, ensuring only intended directories are searchable.
- Agent Scoping: A deny-by-default policy is applied, mapping specific corpora to authorized agents only.
- Template Bootstrapping: Sanitized configuration templates are generated to define backends, corpora, and memory roots.
- Interface Implementation: A stable search wrapper is established to provide a consistent API for agents while hiding indexing complexities.
- Freshness Tracking: The system monitors file signatures and fingerprints to perform selective, incremental updates instead of full re-indexing.
- Validation: Smoke tests verify broad queries, agent-specific denials, and successful refresh paths.
Workspace Local Retrieval Setup
Install the skill globally using npm to access the integrated CLI tools for Openclaw Skills:
npm install -g workspace-local-retrieval
Alternatively, bootstrap a specific workspace directory without a global installation:
npx workspace-local-retrieval install --dest ./retrieval
To verify your environment meets the necessary requirements for local embeddings and indexing, run:
workspace-local-retrieval check
Workspace Local Retrieval Data Schema & Taxonomy
The skill organizes its architecture into a dedicated retrieval/ directory with a structured taxonomy:
| Directory | Purpose | Key Files |
|---|---|---|
config/ |
System definitions | corpora.json, agent_corpora.json, backend.json |
scripts/ |
Functional logic | workspace_search.mjs, build_index.py, refresh_incremental.py |
indexes/ |
Persistent storage | workspace_retrieval.sqlite, vector artifacts |
references/ |
Documentation | privacy-and-boundaries.md, interface-contract.md |
Workspace Local Retrieval Advanced Features
- Deny-by-Default Security: Ensures agents cannot access any data corpus unless explicitly allowlisted.
- Incremental Refresh Logic: Uses file fingerprinting to update only changed documents, significantly reducing resource consumption.
- Multi-Agent Scoping: Supports complex environments where different agents require different levels of data trust.
- OS-Specific Task Planning: Automatically generates installation plans tailored to macOS, Linux, or Windows dependencies.
- Stability Wrappers: Decouples the agent-facing search interface from the underlying embedding backend for easier future migrations.
SKILL.md
Loading
Related Openclaw Skills
Sendoso
A comprehensive integration for managing personalized gifting, eGifts, and direct mail campaigns through AI agents.
ChartGen AI
ChartGen AI is a comprehensive visualization and data analysis skill that transforms raw data into professional charts, dashboards, and PowerPoint reports.
Veo 3.1 Video Generation
A professional-grade video generation skill for creating cinematic, multi-shot AI videos via the NanoPhoto.AI Veo 3.1 API.
IndexNow Pro
A high-performance SEO tool for instant content indexing across major search engines supporting the IndexNow protocol.
Skill Creator
Skill Creator is a specialized meta-skill designed to help developers build, test, and refine new AI agent capabilities through an iterative, data-driven workflow.
Mambo TTS
A high-performance text-to-speech skill for Openclaw providing lively, pre-configured Chinese voice profiles based on Edge TTS.
Featured*
