kadaliao

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A code-checking skill that tries to break code changes, deployments, and completed tasks instead of only checking whether they appear to work.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A coordinator pattern for managing several AI agents, with one agent directing workers that research, implement and verify work. The coordinator combines their results and makes decisions instead of doing every task itself.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A lightweight code-exploration method for finding files, functions, and architecture details with minimal context.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A structured system for storing and recalling an agent’s long-term memories using four categories and frontmatter, which is metadata written at the top of a file.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A task-execution check that looks for an AI agent trying to skip work, simplify a problem, or replace actual verification with something that merely looks correct.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A memory-management skill for OpenClaw agents, which are software agents that can use saved context. It organizes memories, checks whether they are outdated, and watches for excessive growth.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A set of patterns for running multiple agents or task steps safely while tracking dependencies. It includes ideas such as separating reads from writes and marking tasks as blocked by other tasks.

kadaliao/claude-code-skills-collection

Skill Claude CodeCodex

A guide for writing clear instructions to AI sub-agents, which are separate agents given delegated coding tasks. It focuses on providing enough context and asking for completed work rather than just an explanation.