One-click install
npx skills add https://github.com/swingerman/disciplined-agentic-engineering --skill prime-context
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prime-context
Source: https://github.com/swingerman/disciplined-agentic-engineering/tree/main/engineer/skills/prime-context
Command: npx skills add https://github.com/swingerman/disciplined-agentic-engineering --skill prime-context

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It solves the problem of losing track of a feature’s existing charter, decisions, and related code pointers when an agent starts work on a Ready feature it hasn’t touched in the current session.

Core Features & Use Cases

  • Silent working-memory batch loading: Loads feature.md, CHARTER.md, manifest.yml, and prior handoffs (especially *-discuss.md) plus code/design pointers without narrating intermediate artifacts.
  • Fast orientation and alignment: Summarizes the feature’s outcome, scope, autonomy level (including charter cap), and key prior decisions and ADRs to re-synchronize the agent.
  • Single-step verification interview: Asks exactly one question—anything else to load—then stops, producing no artifact.
  • Session-aware prep between phases: Intended as a prep step between feature-init and discover-acs for features the agent has not worked yet.

Quick Start

Run the prompt: "Prime context for feature X so you can load the feature, charter, manifest, and prior handoffs and then ask one question about anything else to load."

Frequently Asked Questions about prime-context

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

FAQPage Schema
How do I reload feature context and prior handoffs for Claude Code before coding?

You reload feature context by running a prime-context prompt that resolves the feature folder via dae_resolve.py, then batch-loads feature.md, CHARTER.md, manifest.yml, and prior handoffs to restore working memory for the agent.

What is the best way to prepare an agent workflow for a Ready feature it has not touched this session?

Preparing an agent workflow requires priming the agent between feature-init and discover-acs phases, batch-loading the feature charter and prior decisions, asking one verification question, and emitting a handoff summary with no artifacts.

How does silent context loading work for spec-driven engineering features?

Silent context loading batch-reads feature.md, CHARTER.md, manifest.yml, and prior discuss handoffs with code pointers without narrating intermediate artifacts, providing an orientation summary to re-synchronize the agent with the feature scope.

When do I need to prime an agent with feature context in an ATDD pipeline?

You need to prime an agent when starting work on a Ready feature it has not handled in the current session, specifically as a prep step between feature-init and discover-acs phases in the pipeline.

Does prime-context batch-load code and design pointers along with the feature charter?

Yes, prime-context batch-loads specified files including feature.md, CHARTER.md, manifest.yml, prior handoffs, and related code and design pointers to provide a complete orientation summary for the agent.

What happens if the feature.md folder or file is missing during context loading?

If the feature.md folder or file is missing, the context loading process rejects the operation after resolving the methodology root and manifest via dae_resolve.py, preventing the agent from proceeding without a valid feature charter.