Evolve agent behavior into auditable Genes and Capsules with Evolver
Turn recurring agent lessons, logs, and strategy tweaks into reviewable GEP evolution artifacts that can be applied across supported coding-agent runtimes.
npx skills add agentskillexchange/skills --skill evolve-agent-behavior-into-auditable-genes-and-capsules-with-evolver
Use Evolver when an agent workflow keeps producing lessons, mistakes, or strategy improvements that should become auditable reusable evolution artifacts instead of informal prompt tweaks. The upstream project describes a GEP-powered self-evolution engine that scans logs and memory, selects Genes or Capsules, emits a GEP prompt artifact, writes EvolutionEvents for audit, and can wire hooks into Cursor, Claude Code, Codex, Kiro, opencode, and OpenClaw.
What this skill actually does
Invoke this instead of manually editing prompts or rules when the job is to review agent history, select a bounded evolution asset, and apply it through a repeatable CLI or hook workflow. The scope boundary is agent self-evolution and prompt-governance operations inside a git-backed project. It is not a generic agent framework, broad EvoMap platform card, or model leaderboard entry.
Inputs and prerequisites: Node.js 18+, Git, npm, a git-initialized project, optional hooks for Cursor, Claude Code, Codex, Kiro, opencode, or OpenClaw.
Setup notes: npm install -g @evomap/evolver, then run evolver inside a git-initialized project. Use evolver –review for human approval or evolver setup-hooks –platform= to integrate with a supported agent runtime.
Source and verification boundary: use https://evomap.ai/wiki as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Multi-Framework workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.