Home

agent-context overview

intel-agency/agent-context is a GitHub template repository that provides the substrate for AI agent development workflows. Each downstream instance is cloned from this template to house a unique application plan and develop it using a structured, agent-driven process.

The repo ships three main things:

1. An operating manual (AGENTS.md) that tells AI agents how to work in any clone of this repo: coding guidelines, validation rules, source control conventions, delegation patterns, and tool usage.
2. A memory and rules system (.agents/) with durable project context (memory.md), subject-specific rules files (one per topic), and agent skills.
3. The gh-issue-tracking-init skill that scaffolds a GitHub issue-based planning hierarchy (plan to epic to story to task) with linked sub-issues, a Projects v2 board, milestones, and labels, all driven by idempotent PowerShell scripts.

All scripts are written in cross-platform PowerShell 7+ and use the GitHub CLI (gh) for API operations. The repo has 100 tracked files across 120 commits, with 190 Pester tests at 93.91% coverage.

Quick links

• Architecture - system architecture and component relationships
• Getting started - prerequisites, setup, testing
• Glossary - project-specific terms
• How to contribute - development workflow and conventions
• gh-issue-tracking-init - the main skill
• Memory and rules system - the .agents/ directory
• Validation pipeline - build, scan, test

---

Generated by Factory