ryanthedevCommunity·13 Agent Skills Included

oberskills

Prompt design, subagent dispatch, skill testing, and human-sounding writing

Designs and reviews prompts, dispatches subagents with cost-aware delegation rules, and builds tested skills through baseline-first evals. Removes robotic writing patterns, hallucinated research links, and unreliable skill triggers that waste tokens and produce weak output. Includes a bundled eval server that validates, benchmarks, and grades skills automatically so every workflow ships verified.

60|3

↗

npx skills add ryanthedev/oberskills --all -g -y

Available:

Tells the AI agent how this plugin's skills, commands, and eval server fit together, and sets the formatting, evidence, and testing rules it must follow when working on or with these skills.

All Skills in This Repository (13)

Pure Emerald Level Indicators

📦 In Repo

ryanthedev

designing-deep-modules

Generate design approaches and evaluate trade-offs for new modules.

maintaining-design-quality

Identifies design degradation risks in code modifications and guides safer, well-documented changes across modules.

simplifying-complexity

Apply define-out, mask, and aggregate gates to simplify complex codebases.

improving-code-clarity

Codify a comments-first workflow for code readability reviews.

verifying-correctness

Verify code correctness across requirements, concurrency, errors, resources, boundaries, and security.

reviewing-module-design

Analyze module designs for complexity, depth, information hiding, and abstraction weaknesses.

strategic-coding

Identify strategic coding problems and enforce measure-first validation before code changes.

optimizing-critical-paths

Enforce a measurement-first workflow for performance optimization with repeatable benchmarks.

obershot

Capture full-screen or active-window screenshots and summarize them with a Haiku subagent.

oberagent

Gate agent dispatches by mandating the oberprompt checklist before Task tool use.

oberprompt

Audit and improve prompts using FIX and REVIEW modes with validation checks.

obercreate

Create and review skills and agent prompts through phase-driven workflows with quality gates.

Community

Advanced

Frequently Asked Questions

FAQPage Schema

How to install oberskills?▼

Run `npx skills add ryanthedev/oberskills --all -g -y` in your terminal to install all skills in this collection globally.

What does oberskills do?▼

It improves how your coding assistant writes prompts, dispatches subagents, builds and tests new skills, researches the web, and edits prose so it sounds human.

How do I make AI writing sound less robotic?▼

The write skill rewrites or reviews your text using measured detection signals like sentence-length variance and banned phrase lists, cutting AI detection rates dramatically.

Does oberskills work with Claude Code and Codex?▼

Yes. It ships manifests for both hosts, and the shared skill bodies stay portable while host-specific mechanics are clearly labeled.

Can I test a skill before shipping it?▼

Yes. The bundled skill-eval server runs baseline comparisons, trigger-rate checks, and graded benchmarks so you only ship skills that measurably help.