Skill Building

Write SKILL.md files with frontmatter, trigger phrases, decision trees and checklists

Generate agent skills from Rust crate and std library documentation

Fix agent plugins that are installed but missing, or skills that never activate

Build marketplace.json catalogs and plugin suites that package skills for install

Create, edit and benchmark agent skills, then run evals to measure them

Improve an existing agent skill using past successes, failures and competitor evidence

Reconcile duplicate and stale agent skills across installs and catalog visibility

Review agent skill quality with design rubrics and batch inventory checks

Find, inspect and install agent skills and plugins from local repositories

Create LLM-first agent skills with valid frontmatter and clear triggers

Audit and upgrade existing agent skills for clarity and quality

Index available agent skills by trigger and path after skill changes

Scaffold and ship plugins that bundle hooks, tools and skills into one package

Scaffold, edit and delete custom agent skills written as SKILL.md files

Discover bundled skills and search and install community skills from skills.sh

Design, validate and review SKILL.md files and test their trigger descriptions

Route each turn to the right Aegis skill before starting work

Test and harden a skill against golden cases using blind subagent runs

Create or revise agent skill files and their trigger descriptions

Record skill feedback and iteratively improve skill evals with the skill-up CLI

Mine your past coding sessions for repeatable workflows and forge them into skills

Create, test, benchmark and optimize agent skills with evals and description tuning

Update installed paperthin skills, retire renamed ones and add missing ones

Create new Claude Code skills, slash commands and plugin skills from a request

Search skills.sh and GitHub repos, then install agent skills you choose

Add or audit copyright comment headers across skill files like SKILL.md and scripts

Explain and score local agent skills for safety, structure and readiness

Create a new Kai skill from one plain-language request and install it

Find and install agent skills from the Orthogonal skill library

Create and package agent skills with SKILL.md files, scripts and references

Create and convert multi-role Teamskills with roles, workflows and bind constraints

Build Claude Code skills, sub-agents, hooks, memory, settings and plugins

Suggest skills to an operator without selecting or loading them automatically

Evolve agent skills from run trajectories and verify changes before applying them

Rank an installed skill catalog to pick the right skill for a request, if any

Install a library of bioinformatics skills for sequencing, RNA-seq and structure tasks

Test four variations of an installed skill and apply the one that scores best

Validate installed skill output against assertion manifests and catch regressions

Diagnose and fix a failing installed skill with the smallest root-cause change

Draft new agent skills with clear triggers, workflows and bundled scripts

Author SKILL.md files with correct frontmatter, validator limits and structure

Analyze usage logs and propose optimizations to skill files, reviewer prompts and defaults

Tune a skill file against an LLM judge rubric until a test agent performs well

Scaffold a new agent skill with SKILL.md, code skeleton and tests from a spec

Guide the creation and updating of agent skills that add specialized workflows

Check a skills library for stale skills, validation errors and out-of-sync docs

Browse a skills catalog and compare candidates before installing any of them

Regenerate README and work-area docs for a managed skills library with a dry run

Build a team skills library by mapping work to shelves and writing notes