self-inspection

Inspect Hexis source code and live PostgreSQL database schema for implementation evidence.

609|81|Updated Nov 15, 2024
One-click install
npx skills add https://github.com/QuixiAI/Hexis --skill self-inspection
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-inspection
Source: https://github.com/QuixiAI/Hexis/tree/main/skills/installed/self-inspection
Command: npx skills add https://github.com/QuixiAI/Hexis --skill self-inspection

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Answering questions about how an agent system is implemented often relies on guesswork or stale documentation. This Skill grounds every implementation claim in direct evidence by reading the actual Hexis source code and querying the live PostgreSQL schema.

Core Features & Use Cases

  • Source Code Inspection: List, search, and read bounded line ranges of repository files to locate definitions and call sites before making claims.
  • Live Schema Inspection: Search tables, columns, views, and stored functions, then describe relations or retrieve function definitions to learn what is actually running in PostgreSQL.
  • Defect Diagnosis: Review recent changes and use self-repair workflows to list, diagnose, and resolve defect reports with source and schema evidence.
  • Use Case: When asked why a belief revision behaves unexpectedly, search the source for the revision policy, inspect the live schema for the relevant stored function, and reconcile any differences between baseline SQL, migrations, and the running database.

Quick Start

Ask the agent to inspect its source code and explain how the heartbeat cycle is implemented, citing the relevant files and functions.

Frequently Asked Questions about self-inspection

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

FAQPage Schema
How do I inspect a live PostgreSQL database schema?

Use the inspect_database_schema tool: start with search to find a table, column, view, or function name fragment, then call describe_relation for columns, constraints, and indexes, or get_function for stored definitions. The tool is metadata-only and does not accept arbitrary SQL.

How to search source code before making implementation claims?

Use inspect_source with action search to find definitions and call sites, then read only the necessary line ranges with the read action. Follow references across modules rather than inferring behavior from a filename or a single function.

What is the difference between baseline SQL files and the live schema?

Baseline SQL files describe a fresh database, migrations evolve existing databases, and the live schema is the authority for what is running now. The Skill reconciles these explicitly and reports when the running schema differs from baseline files or migrations.

Can the schema inspection tool run arbitrary SQL queries?

No. The inspect_database_schema tool is metadata-only and does not accept arbitrary SQL. It supports overview, search, describe_relation, and get_function actions for reading schema structure and stored function definitions.

How do I diagnose failing tool calls or heartbeat steps?

Use the self_repair tool to list unresolved defect reports, diagnose one using source and schema evidence, and mark it resolved only after verification. It records and reasons about defects but never silently edits source code.