greglas75/zuvo · skills/skill-eval/SKILL.md

skill-eval skillA

skill-eval is agent-read markdown (skill) from greglas75/zuvo: Behavioral evaluator for zuvo skills. Runs a skill against its eval corpus (evals/<skill>.evals.json) in fresh executor sub-agents, grades each run transcript against the corpus assertions with an injection-hardened grader, and writes a per-assertion pass/fail report to zuvo/reports/. Supports old-vs-new comparison via --compare <ref>. Dev-only (needs the repo's evals/ + .git)..

Indexed from public GitHub and served as immutable, content-addressed versions. Install it pinned to an exact SHA-256 with the mdr CLI, and every file is verified against the hash recorded here before it reaches your agent. The deterministic audit below grades the latest version, and the same file always earns the same grade.

How to install

Latest version
mdr add greglas75/zuvo/skill-eval@git:20260906.015072f
Exact content
mdr add greglas75/zuvo/skill-eval@sha256:d4e033917f1df53a

Pin to a label to follow the author's releases, or to a sha256 to freeze the exact bytes forever. Either way the resolved hash is written to mdr.lock, and mdr install reproduces it on any machine.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_ab5ooyvlojgbvgp2.svg)](https://markdownregistry.com/a/art_ab5ooyvlojgbvgp2)

0 badge views in 30 days

Versions

versioncommittedcommitsizeaudit
git:20260906.015072f latest2026-09-06 015072f 34,529 BA view · diff
git:20260810.40e315f2026-08-10 40e315f 34,753 BA view · diff
git:20260806.6edffe22026-08-06 6edffe2 34,277 BA view · diff
git:20260802.56ae2272026-08-02 56ae227 33,590 BA view · diff
git:20260710.02267d32026-07-10 02267d3 33,572 BA view

Audit of the latest version

A  17 of 17 checks passed. Deterministic, no model, same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (34529 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

greglas75/zuvo · 6 stars · license MIT · pushed 2026-09-17 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ab5ooyvlojgbvgp2
GET https://markdownregistry.com/api/v1/resolve?ref=greglas75/zuvo/skill-eval
GET https://markdownregistry.com/api/v1/blob/d4e033917f1df53a2a3bfe40467c224557c75331f87e2fb31d575dcac2e79f76

Agents talk at modelranch.com: hand yours the instructions there and it joins the network that reads files like this one.