cap-evolve

A plugin marketplace entry for cap-evolve, a framework for improving AI-agent capabilities against evaluations. It includes skills for intake, implementation, testing, evaluation, diagnosis, comparison, and reporting.

In plain words
What is it for?
Use it to establish baselines, implement and check capability changes, evaluate results, diagnose failures, apply gates, and produce final reports.
Why use it?
It provides an organized workflow for measuring and improving an agent while keeping training, validation, and test results separate. This helps distinguish genuine improvement from overfitting to one evaluation.

Plugin for Claude Code

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Claude Code
/plugin marketplace add skillberry-ai/cap-evolve
agentmods
npx agentmods add plugins/skillberry-ai/cap-evolve/marketplace
Clone the repo
git clone --depth 1 https://github.com/skillberry-ai/cap-evolve

Made for: Claude Code.

Per session not measured What this adds to a session before it is invoked.
When invoked not measured Not applicable: nothing here is loaded into a session.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Security

Grade A, and why

cap-evolve scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude-plugin/marketplace.json · 16 lines

What it actually says

{
  "name": "cap-evolve",
  "owner": { "name": "cap-evolve contributors" },
  "metadata": {
    "description": "Skills-native, host-agnostic harness for optimizing AI-agent capabilities against an eval, with honest train/val/test discipline.",
    "version": "0.1.0"
  },
  "plugins": [
    {
      "name": "cap-evolve",
      "source": "./plugins/cap-evolve",
      "description": "The full cap-evolve skill library: intake, implement-and-check, evaluate, diagnose, gate, baseline, finalize, report; capability/algorithm/optimizer skills; and an autonomous orchestrate skill."
    }
  ]
}
Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 16 lines scan A 4eacf696668a

Subscribe to this mod's changes

cap-evolve is a plugin published in the GitHub repository skillberry-ai/cap-evolve (47 stars, last pushed 2d ago), licensed Apache-2.0. Its token cost is not measured: this kind of file is read by the harness, not the model. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.