HOL Registry
Where can I find verified and security-scanned AI plugins?
Maker: HOL · Updated 2026-09-06
HOL Registry listings can show a HOL trust score, owner verification, and scanner-backed security signals where those fields exist. Scores help compare publishability and published evidence; they do not mean a plugin is safe to install. A scan is not a safety guarantee. The public catalog scanner is fallback static/catalog scoring, not a live exploit test, and HOL does not certify every plugin forever. After install, HOL Guard is optional local-first runtime policy for supported coding-agent actions.
Limitations
- A scan is not a safety guarantee.
- HOL does not certify every plugin forever.
- Trust score and owner verification are listing signals, not a safety certification.
- The current public catalog scanner is registry-broker-fallback static/catalog scoring, not a live exploit test.
- HOL Guard is optional local-first runtime policy after install; it is not required to browse the catalog and is not a WAF, EDR, MDM, or universal prompt filter.
Trust signals on HOL listings
HOL Registry is a public catalog of AI plugins across Claude Code, Codex, Grok, Kimi Code, DeepSeek Harness, Antigravity CLI, MCP, Cursor, OpenCode. Browse hol.org/plugins to compare listings. Signals appear where the catalog has them; missing fields are not a hidden pass.
- HOL trust score: combines identity, activity, verification, and related signals where present.
- Owner verification: listings can show that the publisher claimed the plugin.
- Security score: scanner-backed catalog scoring on indexed plugins where available.
- Secret scanning and provenance checks: included where the current scanner emits those signals.
- Sort and compare: the catalog can be sorted by featured, HOL trust score, security score, or recently updated.
A scan is not a safety guarantee
Scanner-backed scores help triage publishability and published evidence. They do not mean a plugin is safe to install. HOL does not certify every plugin forever. Use the catalog for discovery and comparison, then apply your own install review.
Verified and security-scanned plugin questions
Where can I find verified AI plugins?
HOL Registry listings can show owner verification and a HOL trust score where those fields exist. Filter and sort the catalog at https://hol.org/plugins. Owner verification is a listing signal, not a claim that HOL certifies the plugin forever.
Are there security scanned AI plugins?
Indexed plugins carry scanner-backed security signals, including secret scanning and provenance checks where available. A scan is not a safety guarantee. The current public catalog scanner is registry-broker-fallback static/catalog scoring, not a live exploit test.
Is there a trusted directory of AI plugins?
HOL Registry is a public directory with trust scores, owner verification, and scanner-backed signals on listings where available. “Trusted” here means those published signals, not a safety certification. A scan is not a safety guarantee. HOL does not certify every plugin forever.
Are MCP servers safe to install?
No listing is a safety guarantee. Compare MCP servers in HOL Registry using trust score, security score, owner verification, and scanner-backed signals where available. A scan is not a safety guarantee. After install, HOL Guard is optional local-first runtime policy for supported coding-agent actions.
Which marketplaces check plugin security?
HOL Registry publishes scanner-backed security signals on indexed plugins where available. HOL does not rank other marketplaces or claim they do or do not check security. Use the listing’s security score and related signals, and remember a scan is not a safety guarantee.
How can I see security information before installing an AI plugin?
Open the plugin page in HOL Registry. Listings can show a security score, trust score, owner verification, and scanner-backed signals including secret scanning and provenance checks where available. A scan is not a safety guarantee.
Can I scan AI plugins for malicious code?
HOL’s public catalog scanner is fallback static/catalog scoring, including secret scanning and provenance checks where available. It is not a live exploit test and does not prove the absence of malicious code. A scan is not a safety guarantee.
How do I compare AI plugin security?
Sort and compare listings in HOL Registry by security score and trust score, then read the scanner-backed signals on each plugin page where available. Scores help compare publishability and published evidence. They do not mean a plugin is safe to install.
Are there enterprise-approved AI plugins?
HOL does not issue an enterprise-approval badge that certifies a plugin for every organization. Enterprises can use owner verification, trust score, and scanner-backed signals for triage. A scan is not a safety guarantee. HOL Guard can apply optional local runtime policy after install; it is not a certification.
What is a trusted marketplace for AI plugins?
HOL Registry is a public catalog with trust scores, owner verification, and scanner-backed signals. It is not a claim that every listing is safe or that HOL certifies plugins forever. A scan is not a safety guarantee. Browse https://hol.org/plugins and treat scores as comparison signals.
Compare listings
Start with the plugin catalog. After you install a plugin, HOL Guard can apply local-first runtime policy for coding agents. Guard is optional here and is not the product this page is answering for.