magent-evaluate

magent-evaluate is a command for coding agents from gobing-ai/superskill. It costs 9 tokens per session (667 once invoked), scanned A, original, Apache-2.0.

A command that scores a main coding-agent configuration across five quality areas without changing it.

In plain words
What is it for?
Use it to evaluate a configuration, produce machine-readable results or a scoring work order, save evaluations, or validate an agent-scored result.
Why use it?
It provides a repeatable way to inspect configuration quality and compare results before and after revisions.

Command

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.

agentmods
npx agentmods add commands/gobing-ai/superskill/magent-evaluate
Clone the repo
git clone --depth 1 https://github.com/gobing-ai/superskill

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for magent-evaluate

README.md
[![agentmods](https://agentmods.dev/badge/commands/gobing-ai/superskill/magent-evaluate.svg)](https://agentmods.dev/commands/gobing-ai/superskill/magent-evaluate)
Your own site
<a href="https://agentmods.dev/commands/gobing-ai/superskill/magent-evaluate"><img src="https://agentmods.dev/badge/commands/gobing-ai/superskill/magent-evaluate.svg" alt="Measured on agentmods" height="20"></a>
Per session 9 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 667 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00009 $0.00667
Opus 5 $0.00005 $0.00333
Sonnet 5 $0.00002 $0.00133
Haiku 4.5 $0.00001 $0.00067

Measured today against content hash 66fcccac6511, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

magent-evaluate 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 today.

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.

plugins/cc/commands/magent-evaluate.md · 71 lines

How it starts

The opening of the file, as written. The whole thing — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Magent Evaluate

Wraps cc:cc-magents skill.

Score main-agent config quality across five dimensions, then conduct the semantic audit the heuristics cannot prove. Evaluate only — make NO changes. Delegates to cc:cc-magents skill.

When to Use

  • Check current score without making changes
  • Compare scores before and after refinement
  • Audit a resolved multi-file or multi-target instruction set for drift and contradictions

Arguments

Argument Description Default
<nameOrPath> Configuration name or path to its .md file (required)
--json Output machine-readable JSON; with --rubric, emit a scoring work order false
--target Target platform claude
--save Persist the evaluation to the evaluation store (enables evolve trend analysis) false
--rubric <file> Rubric path for envelope-out scoring built-in
--ingest <file> Agent-scored result JSON to validate and persist -
--base-path <dir> Directory relative markdown links resolve against, so a governance area satisfied by a link to an existing file counts toward completeness. Override when the config is authored to live elsewhere — e.g. scoring a scaffold template as if already at a project root. the file's own directory

Examples

# Evaluate a config
/cc:magent-evaluate ./CLAUDE.md
# Save results to the evaluation store
/cc:magent-evaluate ./CLAUDE.md --save

Implementation

Pass $ARGUMENTS to the underlying skill for processing.

Delegates to cc:cc-magents skill:

Skill(skill="cc:cc-magents", args="evaluate $ARGUMENTS")

Direct CLI execution (all platforms):

superskill magent evaluate $ARGUMENTS

Evaluation Contract

Follow cc:cc-magents workflows § "Main-Agent Evaluation and Refinement": resolve the actual load graph, run strict validation, record every baseline dimension, and report heuristic findings separately from semantic findings. Check authority and scope, stable/non-inferable content, live links, exact leaf-command help, the reasoned tool ladder, safety boundaries, and runnable gates. An aggregate score or exit code alone is not a PASS.

Read the full file on GitHub · 71 lines

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. today Changed · +13 lines 66fcccac6511
  2. 3d ago First seen · 58 lines · 9 tokens per session scan A 5f931329ef17

Subscribe to this mod's changes

magent-evaluate is a command published in the GitHub repository gobing-ai/superskill (5 stars, last pushed yesterday), licensed Apache-2.0. It adds 9 tokens to every session and 667 once invoked, about $0.0000 per session on Opus 5. 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 2026-08-31.