improve
The improve skill acts as a senior advisor that audits codebases to identify improvement opportunities and generates self-contained implementation plans for other agents to execute. It performs read-only analysis, including recon, auditing, and plan creation, while strictly avoiding direct code modification.
Is improve safe to install?
Review the source first: our audit of improve's source files found 8 shell commands, 0 external URLs, file reads and writes (high risk). Every command and URL listed appears verbatim in the skill's source. The skill reads local files and executes read-only analysis commands. It can create or modify files in the plans/ or advisor-plans/ directories. It may execute shell commands for verification and interact with GitHub via the gh CLI if explicitly requested.
How we audit skills: our security review methodology.
Who is this skill for?
Developers and teams who want to audit their codebase for bugs, security, performance, or technical debt and require structured, actionable implementation plans for other AI agents to execute.
What can you do with it?
- Auditing a codebase for bugs, security vulnerabilities, and performance issues.
- Generating prioritized implementation plans for technical debt, migrations, or new features.
- Creating self-contained handoff plans for other AI agents to execute.
- Reviewing existing plans for quality and clarity.
- Executing plans via a subagent and reviewing the resulting diffs.
How good is this skill?
Quality score: 5/10. The skill documentation is comprehensive, clearly defining its role as an advisor rather than an implementer. It includes strict safety rules, a defined workflow, and specific invocation variants.
What does the skill file contain?
# Improve You are a **senior advisor, not an implementer**. Your job is to deeply understand a codebase, find the highest-value improvement opportunities, and write implementation plans good enough that a *different, less capable model with zero context from this session* can execute, test, and maintain them. The economics of this skill: an expensive, high-ceiling model does the part where intelligence compounds (understanding, judging, specifying). Cheaper models do the execution. The plan is the product — its quality determines whether the executor succeeds. ## Hard Rules 1. **Never modi...
Frequently asked questions
Does this skill modify my source code?
No. The skill is strictly read-only regarding source code. It only creates or modifies files within the plans/ or advisor-plans/ directories.
How does the skill handle sensitive information like credentials?
The skill never reproduces secret values. If it finds credentials or tokens, it references the file and line number only and recommends rotation.
Can I use this to implement features directly?
No. If asked to implement directly, the skill declines and offers to create a plan or dispatch an executor subagent for review.
What happens if the repository has no tests?
The skill records this during the recon phase and often prioritizes establishing a verification baseline as a primary finding.
Related skills
find-skills
2.3MUsers seeking to extend agent capabilities with specialized tools, workflows, or knowledge packages
The find-skills skill enables agents to search for, discover, and install modular packages from the open agent skills ecosystem using the Skills CLI.
video-edit
338.7KUsers of the RunComfy CLI who need to automate video editing tasks like restyling, background swapping, or motion transfer
The video-edit skill acts as a router for the RunComfy CLI, selecting between Wan 2.7 Edit-Video, Kling 2.6 Pro Motion Control, and Lucy Edit Restyle models based on user intent to perform video transformations.
lark-doc
305.2KUsers who need to automate document management, content updates, and media handling within the Lark/Feishu ecosystem
The lark-doc skill enables agents to read, create, and edit Lark (Feishu) documents, including Docx and Wiki formats. It supports content manipulation via XML or Markdown, media handling, and resource management for document covers. The skill integrates with other Lark skills by identifying and delegating operations for embedded objects like spreadsheets, databases, and mind notes.
brainstorming
256.4KDevelopers and project managers who need to define requirements, architecture, and design for software projects before writing code
A structured brainstorming framework that guides users through project exploration, design proposal, and specification before implementation. It enforces a strict design-first workflow, requiring user approval for designs and documentation before transitioning to the writing-plans skill.