knowledge

Manage Knowledge DB entries for modules via index.yaml and git semantics.

Updated Apr 14, 2026
One-click install
npx skills add https://github.com/sr-ai-dev/sr-harness --skill knowledge-sr-ai-dev
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: knowledge
Source: https://github.com/sr-ai-dev/sr-harness/tree/main/skills/knowledge
Command: npx skills add https://github.com/sr-ai-dev/sr-harness --skill knowledge-sr-ai-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the problem of missing or outdated, module-specific knowledge that causes inefficient and error-prone development during requirement discovery and context research.

Core Features & Use Cases

  • Module Knowledge DB CRUD: Scan, update, delete, list, and check status for a structured Syscon Robotics Knowledge DB.
  • Index-driven knowledge management: Maintains a centralized index.yaml so each module’s common and ROS-variant knowledge files are discoverable and verifiable.
  • Re-scan with safe learning preservation: Refreshes knowledge by re-scanning the workspace while preserving accumulated learnings to avoid losing prior insights.
  • ROS-aware knowledge branching: Generates ROS1 or ROS2 specific knowledge files based on detected dependencies (and maintains topic/service/QoS-related data when applicable).
  • Use Case: When working in a brownfield repo, generate or refresh knowledge for a specific module so downstream specify/execution phases have accurate context instead of relying on guesses.

Quick Start

Run /knowledge scan core-driver to generate or update the knowledge DB for the core-driver module from your current workspace.

Frequently Asked Questions about knowledge

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I keep module knowledge bases fresh when working in a brownfield repository?

Keeping module knowledge bases fresh in a brownfield repository requires re-scanning the workspace to update cached knowledge while preserving accumulated learnings. This ensures downstream execution phases have accurate, verified context instead of relying on outdated guesses.

What is the best way to manage ROS1 and ROS2 specific knowledge files?

Managing ROS1 and ROS2 specific knowledge files is handled by generating ROS-variant specific knowledge based on detected dependencies. The system maintains topic, service, and QoS-related data within a centralized index.yaml so each module's files stay discoverable and verifiable.

How does git commit tracking work for knowledge base validation?

Git commit tracking validates knowledge base state by enforcing index.yaml schema and checking filesystem presence alongside commit semantics. This mechanism ensures the knowledge DB state remains consistent and fully traceable across module updates and deletions.

Can I refresh workspace knowledge without losing prior context research?

You can refresh workspace knowledge without losing prior context research by using the re-scan functionality. It refreshes knowledge files tied to a normalized index while safely preserving accumulated learnings to avoid losing previously captured module insights.

When do I need to normalize index.yaml for module-specific knowledge DB entries?

You need to normalize index.yaml for module-specific knowledge DB entries when managing structured knowledge during requirement discovery. It provides a centralized schema so common and ROS-variant knowledge files are discoverable, verifiable, and consistent across the workspace.