skills.rest
Asygnuz-S-A-S
Asygnuz-S-A-SOFFICIAL
Explore all skills by this author →
GitHub

agent-dx-cli-scale

Scores CLI tools against seven agent-first design principles on a 0-21 scale.

Updated May 22, 2026
Search Tags:#developer tools#cli#ai agents#evaluation#scoring rubric#cli design#agent dx
One-click install
npx skills add https://github.com/Asygnuz-S-A-S/one-star --skill agent-dx-cli-scale-asygnuz-s-a-s
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-dx-cli-scale
Source: https://github.com/Asygnuz-S-A-S/one-star/tree/main/.agents/skills/agent-dx-cli-scale
Command: npx skills add https://github.com/Asygnuz-S-A-S/one-star --skill agent-dx-cli-scale-asygnuz-s-a-s

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams building command-line tools lack a structured way to measure how well their CLI works for AI agents, leading to unpredictable parsing, hallucinated inputs, and missing safety rails when agents operate the tool. ## Core Features & Use Cases - Seven-Axis Scoring Rubric: Evaluates machine-readable output, raw payload input, schema introspection, context window discipline, input hardening, safety rails, and agent knowledge packaging, each scored 0-3. - Rating Interpretation Bands: Maps totals of 0-21 to Human-only, Agent-tolerant, Agent-ready, or Agent-first classifications. - Multi-Surface Readiness Checklist: Notes whether the CLI exposes MCP over stdio, plugin installation, and headless authentication. - Use Case: A platform team evaluating their internal deployment CLI runs this rubric, discovers it scores 8 (Agent-tolerant) due to missing dry-run support and no JSON schema introspection, and prioritizes those gaps in the next sprint. ## Quick Start Use the agent-dx-cli-scale skill to evaluate our deployment CLI against the seven agent-first design axes and produce a scored report.

Frequently Asked Questions about agent-dx-cli-scale

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

FAQPage Schema
How do I evaluate whether a CLI works well with AI agents?▼

Score the CLI on seven axes from 0 to 3: machine-readable output, raw payload input, schema introspection, context window discipline, input hardening, safety rails, and agent knowledge packaging. Sum the scores for a total between 0 and 21.

What makes a CLI agent-first instead of human-first?▼

Agent-first CLIs default to structured JSON output, accept raw API payloads, expose runtime schema introspection, harden inputs against hallucination patterns like path traversal, and provide dry-run modes. Human-first CLIs optimize for discoverability and forgiveness instead.

What score ranges does the agent DX scale define?▼

Totals of 0-5 are Human-only, 6-10 are Agent-tolerant, 11-15 are Agent-ready, and 16-21 are Agent-first. Each band describes how much prompt engineering and error compensation agents need to operate the CLI.

Why does input hardening matter for agent-operated CLIs?▼

Agents fail through hallucinations rather than typos, producing path traversals, percent-encoded segments, and embedded query params in resource IDs. Hardened CLIs reject these patterns and sandbox output paths to the current working directory.

Does the scale cover MCP and headless authentication?▼

Yes, as an unscored bonus checklist. It notes whether the CLI exposes MCP over stdio JSON-RPC, supports extension or plugin installation, and allows headless authentication via environment variables without browser redirects.

Related Skills

Explore More Software Engineering Skills→
yspreenyspreen
33

asc-cli-usage

Guide asc CLI command design for App Store Connect with flags, output formats, and authentication.

Community
Intermediate
albanistrefialbanistrefi
4

umbraco-cli

Inspect and mutate Umbraco Management API resources with JSON and dry-run options.

Community
Intermediate
aviskaaraviskaar
6

ai-security-analyst

Analyze AI and ML systems for threats, prompt injection, and hallucination risks.

Official
Advanced
phuhao00phuhao00
12

agent-eval

Benchmarks AI agent performance by comparing tool-assisted navigation against baseline grep and read operations.

Community
Advanced