KnowTrust-Graph for Openclaw

KnowTrust-Graph is an agent-agnostic tool that combines knowledge management with a dynamic verification engine to track the reliability of entities and facts.

ronie-shi
v1.2.0
May 12, 2026
1
301
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install knowtrust-graph

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 knowtrust-graph 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 KnowTrust-Graph?

KnowTrust-Graph is a sophisticated knowledge management framework designed for Openclaw Skills that bridges the gap between raw data and verifiable truth. It operates on a three-layer model: Entities, Relations, and Facts, ensuring that every piece of information is not just recorded but also qualified by its source and verification level.

The skill is built on a philosophy of Human-led, AI-assisted interaction, where the AI organizes and suggests connections while humans maintain decision-making authority. It is completely lightweight, requiring only Python and SQLite, making it ideal for developers and small teams who need a robust, local-first knowledge base without the overhead of complex vector databases or external cloud dependencies.

KnowTrust-Graph Use Cases

  • Managing personal project knowledge, learning notes, and decision logs.
  • Enabling small team information sharing (3-10 people) with branch support.
  • Tracking the verification status and expiration of critical business constraints.
  • Automating structured knowledge extraction from long chat dialogues or technical documentation.

How KnowTrust-Graph Works

  1. Entity & Fact Definition: Users or AI agents define entities (people, projects, concepts) and attach atomic facts to them.
  2. Trust Chain Attribution: Every fact is linked to a source and assigned an initial verification state (unverified).
  3. Dynamic Verification: As facts are checked, their state evolves from self_checked to peer_reviewed or external_confirmed based on team consensus.
  4. Lifecycle Management: The system tracks the freshness of data, marking old information as outdated or stale to trigger updates.
  5. Query & Exploration: Users search or traverse the graph using CLI commands to retrieve reliable, verified knowledge.

KnowTrust-Graph Setup

Ensure you have Python 3.8+ installed. The skill uses standard libraries and SQLite.

# Create the knowledge directory
mkdir -p ~/.zhixin

# Initialize the database
python3 ~/.zhixin/kg.py init

# Check status
python3 ~/.zhixin/kg.py stats

KnowTrust-Graph Data Schema & Taxonomy

KnowTrust-Graph organizes data into a relational structure optimized for traceability:

Component Description Metadata Examples
Entity Core nodes in the graph Type (person, project, tool), Description
Relation Directed edges between entities Type (works_on, depends_on, created_by)
Fact Atomic knowledge statements Content, Fact Type, Freshness Date
Source The origin of the knowledge Reliability Rating, Author
Annotation Team feedback/validation Type (confirm, dispute, clarify)

KnowTrust-Graph Advanced Features

  • Multi-Agent Protocol: Standardized CLI interface compatible with Openclaw Skills, Claude Code, and Continue.dev.
  • Semantic Search: Optional integration with Ollama for local vector-based embedding and semantic retrieval.
  • Branching System: Create isolated knowledge branches for different projects or experimental data.
  • Consensus Engine: Automatic calculation of Verifiability scores based on source reliability, verification depth, and consensus ratio.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*