capability-assess-ai

Assess AI development tools and record decisions in tech-stack.md.

Updated Mar 31, 2026
One-click install
npx skills add https://github.com/learnn-com/engineering-kb --skill capability-assess-ai-learnn-com
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: capability-assess-ai
Source: https://github.com/learnn-com/engineering-kb/tree/main/.skills/capability-assess-ai
Command: npx skills add https://github.com/learnn-com/engineering-kb --skill capability-assess-ai-learnn-com

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Choosing and documenting AI development tools (assistants, MCP integrations, AI SDKs) is often ad-hoc and inconsistent across teams. This Skill provides a structured resolution cascade to evaluate AI tooling against guidelines, write the AI section of tech-stack.md, and record the decision as an ADL entry. ## Core Features & Use Cases - Resolution Cascade: Resolves decisions via argument override, existing adoption, or full assessment, avoiding redundant work. - Guideline-Driven Evaluation: Reads AI development guidelines and project context (PRD, existing stack) to recommend tools with versions and a maturity level target. - Adoption & Decision Recording: Writes only the AI section of tech-stack.md while preserving other sections, then composes /capability-record-decision to log the decision. - Use Case: During project bootstrap, run the assessment to select a primary AI assistant (e.g., Claude Code or Cursor), decide on MCP adoption, and persist the choice so the whole team shares one documented decision. ## Quick Start Run the capability-assess-ai skill to evaluate AI development tools for this project and write the AI section of tech-stack.md.

Frequently Asked Questions about capability-assess-ai

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

FAQPage Schema
How do I choose AI development tools for a project?▼

Use a structured assessment that reads AI development guidelines and project context, evaluates tools on code generation, context awareness, and team integration, then records the decision. This skill writes the chosen tools with versions into the AI section of tech-stack.md.

How to document AI tool decisions in a tech stack file?▼

Write the AI assistants, MCP integrations, and AI SDKs with versions into a dedicated AI section of tech-stack.md, preserving other sections. Then record the decision as a non-architectural ADL entry via a decision-recording skill.

What is the resolution cascade for technology assessment?▼

The resolution cascade checks three paths in order: an explicit argument override, an existing adoption entry, or a full assessment. It makes the skill idempotent so re-running it confirms existing decisions instead of duplicating work.

Can I skip the AI tool assessment with a predefined choice?▼

Yes, pass the choice argument (e.g., claude-code or cursor) to skip the full assessment. The skill confirms the choice, checks for conflicts with existing adoption, and proceeds directly to writing the adoption file.

What happens if tech-stack.md already exists without an AI section?▼

The skill adds only the AI section while preserving all other content such as core and testing sections. If the file does not exist, it creates one and warns that other sections should be populated by their owning skills.