By ievo-ai
iEvo — self-evolving plugin for Claude Code. Capture lessons, patch local agents and skills, replay logs on upstream updates.
Uninstall iEvo from the current project. Asks the user before removing the project-rules marker block from CLAUDE.md/AGENTS.md. Preserves `.ievo/` evolution logs.
Replay all evolution logs against fresh upstream agents/skills. Run after the iEvo plugin or any of its dependencies have been updated.
Capture a lesson learned and integrate it into the right place — a local agent file, a local skill file, or the project's CLAUDE.md/AGENTS.md (for project-wide rules). Use when the user identifies a behavior to improve, a mistake to prevent, a project convention, a team role, a tech-stack constraint, or any pattern worth persisting beyond the current session. Patches the target and appends a section to the appropriate `.ievo/evolution/` log.
Submit feedback about the iEvo plugin — bug reports, feature requests, suggestions, or general comments. Posts as a GitHub issue in `ievo-ai/skills` via `gh` CLI. Use when the user says "send feedback", "report a bug", "this didn't work", "I want to suggest a feature", "where do I file an issue for iEvo", or after iEvo has done something the user would want to comment on.
Initialize iEvo in the current project — detect stack and dependencies, discover relevant agent skills from skills.sh through an interview, and prepare the project's evolution log structure. Use when the user runs `/ievo:init`, opens a new project that does not yet have `.ievo/`, or asks "set up iEvo here", "initialize iEvo", or "find skills for this project".
Uses power tools
Uses Bash, Write, or Edit tools
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Capture lessons, patch local agents and skills, replay logs on upstream updates.
iEvo turns ad-hoc feedback ("the agent forgot to check git status before commits") and project context ("we use Unity, our team uses trunk-based dev") into persistent improvements — recorded in markdown evolution logs that survive upstream plugin updates.
iEvo ships as a Claude Code plugin marketplace.
/plugin marketplace add ievo-ai/skills
/plugin install ievo@ievo-skills
/reload-plugins
Then in your project:
/ievo:init
init is an interactive skill that uses find-skills to discover skills relevant to your project and walks you through installing them one at a time.
To make iEvo's marketplace and plugin available only in this project (not globally for your user), add to your project's .claude/settings.json:
{
"extraKnownMarketplaces": {
"ievo-skills": {
"source": {
"source": "github",
"repo": "ievo-ai/skills"
}
}
},
"enabledPlugins": {
"ievo@ievo-skills": true
}
}
Skills (slash-invocable AND auto-activatable, cross-platform via agentskills.io):
| Skill | What it does |
|---|---|
/ievo:init | Discover and install relevant skills from skills.sh. Deep manifest scan (30+ stacks), per-skill interview. Sets up .ievo/evolution/ structure. |
/ievo:evolution "<lesson>" | Capture a lesson. Routes to the right place: an agent, a skill, or the project's CLAUDE.md. Auto-activates when the user expresses a lesson worth persisting. |
/ievo:feedback | Submit a bug report, feature request, or general feedback. Posts a public GitHub issue to ievo-ai/skills via gh CLI. Auto-activates when the user expresses dissatisfaction or wants to suggest something. |
Commands (explicit-only invocation, Claude Code-specific):
| Command | What it does |
|---|---|
/ievo:uninstall | Ask the user, then remove iEvo's marker block from CLAUDE.md / AGENTS.md. Preserves all of .ievo/. |
/ievo:update | Replay evolution logs against fresh upstream agents/skills (after /plugin update). |
Why the split: project-setup operations should be strictly explicit — we don't want the model deciding to remove plugin state on its own. Discovery and lesson capture are agentic — the model picking up "we should remember X" or "this project would benefit from skill Y" is a feature, not a bug.
iEvo does not inject anything into your CLAUDE.md / AGENTS.md at install time. The first time you record a project-wide lesson via /ievo:evolution (e.g. "we use Unity", "PR titles start with the task ID"), the evolution skill:
.ievo/evolution/project.md and appends the rule as a dated section.CLAUDE.md (or AGENTS.md if no CLAUDE.md):
<!-- ievo:start -->
@.ievo/evolution/project.md
<!-- ievo:end -->
After that, every Claude Code session in your project automatically loads project.md because of the @ import in CLAUDE.md. Subsequent project-wide lessons just append to project.md — no further CLAUDE.md changes.
If you /ievo:uninstall, the marker block is removed (after confirmation). The project.md and all evolution logs are preserved.
After /ievo:init:
<your-project>/
└── .ievo/
└── evolution/
├── agents/ # (empty) per-agent logs go here
└── skills/ # (empty) per-skill logs go here
After first project-wide evolution:
<your-project>/
├── CLAUDE.md # ← marker block injected, references .ievo/evolution/project.md
└── .ievo/
└── evolution/
├── project.md # ← project-wide rules log
├── agents/
└── skills/
After init runs (diagnostic logs):
<your-project>/
└── .ievo/
└── log/
├── init-20260517-143200.md # ← each /ievo:init writes one
└── init-20260517-150412.md
Logs capture: detected stack + dependencies, the exact prompt sent to find-skills and its raw response, dedup outcomes (which skills got dropped and why), your install/skip choices, install results per skill. Useful when iEvo suggested something off and you want to file a precise bug — /ievo:feedback can attach the latest log automatically.
.ievo/log/ is intended for local debugging, not version control. Init auto-adds it to your .gitignore if you have one.
npx claudepluginhub ievo-ai/skills --plugin ievoComprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
Write feature specs, plan roadmaps, and synthesize user research faster. Keep stakeholders updated and stay ahead of the competitive landscape.
Curate auto-memory, promote learnings to CLAUDE.md and rules, extract proven patterns into reusable skills.