One-click install
npx skills add https://github.com/lciacci/tessera --skill icpg-lciacci
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: icpg
Source: https://github.com/lciacci/tessera/tree/main/skills/icpg
Command: npx skills add https://github.com/lciacci/tessera --skill icpg-lciacci

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

iCPG prevents developers and autonomous agents from changing code without understanding its original purpose, constraints, ownership, and risk. It reduces duplicated effort and exposes when implementation drifts from the intent and contracts that created it.

Core Features & Use Cases

  • Intent Graph: Create and manage ReasonNodes that connect goals, decisions, owners, contracts, and code symbols.
  • Pre-Task Analysis: Query prior work, file constraints, and symbol risk before making changes.
  • Drift Detection: Compare recorded symbols and contract predicates against the current codebase to identify changed or invalidated intent.
  • Workflow Bookkeeping: Record modified symbols, inspect coverage, bootstrap intents from Git history, and adjudicate drift events.
  • Use Case: Before modifying an authentication service, query for prior authentication work, inspect its constraints and risk profile, then record the affected symbols and check whether their contracts still hold.

Quick Start

Initialize the iCPG graph, create an intent for your planned change, run the three pre-task queries, record modified symbols, and check for drift before fulfilling the intent.

Frequently Asked Questions about icpg

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

FAQPage Schema
What is architectural drift detection in software engineering?

Architectural drift detection compares recorded code symbols and contract predicates against the current codebase to identify when implementation has deviated from the original intent and constraints. It exposes invalidated purpose and ownership risk before changes are merged.

How do I track code intent before modifying a service?

You can track code intent by creating ReasonNodes in an intent-augmented code property graph. Query prior work, inspect file constraints and symbol risk profiles, then record modified symbols to check whether existing contracts still hold.

Can I bootstrap an intent graph from existing Git history?

Yes, you can bootstrap intents from Git history to populate the intent graph. This optional workflow records prior work and constraints into a per-project SQLite graph, establishing a baseline for future drift analysis and contract validation.

How does a code property graph prevent duplicate development work?

A code property graph prevents duplicate work by connecting goals, decisions, owners, and contracts to specific code symbols. Before starting a task, developers query prior work to understand existing constraints and avoid overlapping efforts.

Do I need a specific database to run drift analysis on my codebase?

You need a per-project SQLite graph to run drift analysis. This local database stores the intent-augmented code property graph, including symbol checksums, contract predicates, and ReasonNodes required for pre-task queries and drift detection.

When should I run contract validation during autonomous development?

You should run contract validation before and after code changes. Pre-task queries inspect constraints and risk profiles, while post-change drift detection checks whether recorded contract predicates still hold against the modified symbols.