app-evolver
Agent skill fromdaymade/claude-code-skills
Improves an existing application using its code, product contract, authorized work history and competitor evidence. Use for 取长补短、取其精华去其糟粕、参考竞品优化应用 or evolving an app from successful and failed work. Carries authorized changes through implementation and user-result verification, including REA and reverse Skill composition; use skill-evolver for Skills.
How to install app-evolver
install app-evolver
# install the claude-code-skills pack with the skills CLI
npx skills add daymade/claude-code-skills# or add only this skill to Claude Code: copy the folder
app-evolver/
# into
~/.claude/skills/app-evolver/ (all projects)
.claude/skills/app-evolver/ (one project)
1# App Evolver23Improve the application's actual user result. Extract transferable mechanisms,4resolve their fit with the current product, implement authorized improvements,5and verify them through the application's consuming surface.67Use this as an orchestration Skill. Keep repository engineering, research,8history access and release operations with their existing owners. For a Skill9target use `skill-evolver`; for competitor intelligence alone use10`competitors-analysis`; for a new product specification use the available11requirements workflow. Do not replace those owners with another execution engine.1213## 1. Bind the application and unfinished result1415Before research or edits, establish:1617- Exact application, repository/ref, platform and runtime or deployment identity.18- Original user request, current product contract and active stage; distinguish19 long-term positioning from the work selected now.20- User-visible improvement, falsifiable acceptance evidence, authorized21 deliverables, protected behavior/data and stopping condition.22- Analysis-only versus implementation authorization, and any unresolved23 product choice whose answer changes the outcome.24- Scope of the learning task. For a broad product or competitor learning25 effort, let the bound device and operator roles and the complete intended26 user result, with its impact judgments, constrain the research questions27 before selecting any single technical scenario for study. Keep a task about28…
Excerpt shown under the repository's MIT license.
