Certification for agent artifacts
Certify the skills your AI agents run.
Neutral certificates for every skill, MCP server and ruleset your agents load. Each one is signed, recorded in a public transparency log, and checkable by anyone, on any agent platform. Don't trust us. Verify it.
Signed with Sigstore · logged in public Rekor · built on AGENTS.md, MCP, A2A, SKILL.md
The problem
Your agents already run code nobody reviewed.
Artifacts are executable
A skill or MCP server runs with the agent's full access — files, credentials, memory, network. The instructions are the code.
Registries don't vet
Publishing to a public catalog is a record, not a review. Listing says nothing about whether an artifact is safe to run.
Signing proves who, not safe
A signature confirms the publisher. It doesn't catch prompt injection buried in a skill — and that's already widespread.
How it works
Certify, verify, gate — on the formats you already use.
Certify
Open-source certifier code checks each artifact for provenance and policy conformance, scans it for known injection patterns, and runs a reproducible test suite. The result is a signed certificate, or an A–E trust band where no deterministic test exists.
Verify
Every certificate and revocation goes into a public, append-only transparency log. Anyone can check it, and anyone can re-run the tests and get the same result. You don't need to trust workrr.
Gate & audit
Your agents resolve against a private registry of certified, version-pinned artifacts. Nothing uncertified loads, revocation is instant, and every decision is on record for your auditors.
Verifiable by design
Trust you can check, not trust you have to give.
The certifier is public code
Certificates are signed by an open-source workflow pinned to an exact commit. You can read exactly what produced every result. No hidden allowlist.
No single party in control
Logged in Sigstore's public Rekor, run by the open-source community. Reproduced certificates need two independent forges to agree, and the record is mirrored across GitHub, GitLab and Codeberg.
We never see your code
For internal artifacts, the certifier runs in your own CI. Only a hash goes in the public log. That proves the certificate exists, and when, without revealing anything else.
Plans
Free in public. Paid in private.
Public
Free
Certify open-source skills and MCP servers, verify any certificate with the open CLI, and browse the public ledger.
- Public certificates & trust bands
- Open spec and verifier
- Public transparency log
Publisher
Per publisher
For companies shipping MCP servers, skills or agent integrations. Earn the badge enterprises filter on, renewed on every release.
- Reproduced certification
- Recertify on release
- Verified publisher badge
Private · Team & Enterprise
Talk to us
For organizations running agents on internal and third-party artifacts. Certify in your own CI and gate what loads.
- Private certificates & registry
- Policy authoring & audit trail
- Hands-on rollout with our engineers
Get started
Start free today. Talk to us for private artifacts.
Certifying public skills and MCP servers is free and open to everyone: no account, no sign-up. Teams that need private artifacts, a private registry or audit evidence work with us directly.
Use it now: public artifacts
Add one workflow to your public GitHub repo and push. You get a signed, publicly logged certificate in about five minutes.
jobs:
certify:
permissions: { contents: read, id-token: write, attestations: write }
uses: workrr-trust/certify/.github/workflows/certify.yml@v0.1.11
with: { path: skills/my-skill }Enterprise & private artifacts
Certification for internal artifacts in your own CI, a private vetted registry, policy authoring and audit evidence. Book 20 minutes with us.
Book an enterprise briefingOr leave your details and we'll reach out:
Thanks, we'll be in touch.
We'll reply from info@workrr.ai.