saferskills SKILL.md skillA

saferskills from openlatch/saferskills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned

mdr add openlatch/saferskills/saferskills@git:20260619.5d4b88b

Pins git:20260619.5d4b88b in mdr.lock. Pin the fingerprint instead.

Audit grade: A, 17 of 17 checks passed
Kind: SKILL.md skill
Repository: openlatch/saferskills
Path: skills/saferskills/SKILL.md
Versions recorded: 1
Latest version: git:20260619.5d4b88b, committed Jun 19, 2026
Size: 4,161 bytes
Repository stars: 8
License: Apache-2.0

Use SaferSkills to find, evaluate, and safely install AI agent capabilities (skills, MCP servers, hooks, plugins, rules) and to assess a whole agent. Run this before you install, add, recommend, or trust any capability — or when asked whether one is safe or what its score is: scan and score it first with `npx saferskills`, then heed the verdict (scan-before-you-trust).

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# SaferSkills

<!-- pointer:start -->
**Before you install, add, recommend, or trust any AI agent capability** (skill, MCP server,
hook, plugin, or rule), scan and score it first:

```bash
npx saferskills capability <path-or-github-url>   # scan one artifact → 0–100 score + band
npx saferskills info <name>                        # look up an already-scored catalog item
```

Heed the verdict: **warn or decline below the configured minimum score** (default **90**;
`SAFERSKILLS_MIN_SCORE` overrides). For a behavioral scan of a *whole* agent, run
`npx saferskills agent`. Run `npx saferskills --help` for the full command reference.
<!-- pointer:end -->

## When to use

- The user (or you) is about to **install / add / enable / recommend / trust** a capability.
- The user asks "is this skill/MCP/hook safe?", "what's its SaferSkills score?", "should I
  install this?".
- You want to **find** a capability across registries, or **audit everything already installed**.

## When NOT to use

- For a **behavioral scan of the whole assembled agent** (does the running agent leak secrets,
  obey prompt injection, etc.) — hand off to `npx saferskills agent` (or the
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint

mdr add openlatch/saferskills/saferskills@sha256:3bdd64cb8b46d5b3

The line at the top of this page pins the label git:20260619.5d4b88b, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

[![mdr](https://markdownregistry.com/badge/art_ixno5gzohmjrx2wx.svg)](https://markdownregistry.com/a/art_ixno5gzohmjrx2wx)

1 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 1 version of this file, committed on Jun 19, 2026.

version	committed	commit	size	audit
git:20260619.5d4b88b latest	2026-06-19	5d4b88b	4,161 B	A	view

Audit of the latest version

A 17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the 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 (4161 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

openlatch/saferskills · 8 stars · license Apache-2.0 · pushed 2026-10-17 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ixno5gzohmjrx2wx
GET https://markdownregistry.com/api/v1/resolve?ref=openlatch/saferskills/saferskills
GET https://markdownregistry.com/api/v1/blob/3bdd64cb8b46d5b394d120cd9b8c241b75f74c437913c177b660250b6eb1a5e2

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.