Risk block

Skill aisflows/skill-cleaner/skills/risk-block

Clean messy AI-agent skill folders into a usable SKILL.md library.

Install

npx -y skills add aisflows/skill-cleaner --skill risk-block

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use as step 2 of AIS FLOWS Skill Library Cleaner to separate risky AI-agent skill/source items before cleanup, producing RISK_BLOCKLIST.md and RISK_BLOCKLIST.json with install/runtime/API/credential/browser/package-script blockers.

SKILL.md

2.2 KB, 444 tokens by cl100k_base, as published. Nobody here has run it

Skill Library Cleaner / Risk Block

Role

Separate items that must not enter the active clean library yet.

Input:

machine/INVENTORY.json

Output:

reports/RISK_BLOCKLIST.md
machine/RISK_BLOCKLIST.json

Block Signals

Block or quarantine from active shelf when an item includes:

installer commands;
npx, npm install, pip install, shell-pipe installers;
package scripts or hooks;
unknown binaries;
browser profile/cookie workflows;
token/keyring/credential/.env references;
provider/API/billing flows;
unclear license/source inheritance;
destructive filesystem instructions.

Risk routing is not a malware accusation. It is a release-control decision: the cleaner keeps high-impact skills out of the active shelf until a human explicitly approves them.

Decisions

Allowed risk decisions:

active-eligible
blocked-risk
reference-only
needs-human-review
license-review

Routing Matrix

signal	default route	note
install command, package script, binary, shell-pipe installer	`blocked-risk`	never active by default
secrets, cookies, browser profile, keyring, token handling	`blocked-risk`	do not inspect secret material
provider/API/billing/key setup	`needs-human-review` or `blocked-risk`	block when it asks to run/setup; review when it only documents safe usage
browser/live-app/workstation control	`needs-human-review`	may be safe but needs scope approval
unclear source/license	`license-review`	keep out of publishable package
useful but too broad/noisy	`reference-only`	preserve value without making it active

Required Fields

For every non-active item include:

risk route;
severity: low, medium, high, or critical;
concrete evidence;
why it cannot be active now;
what approval or rewrite would clear it.

Pass Condition

No risky item can become canonical or active before a separate review clears it.