saascode
education & learning·run 237 · Jun 2026

GuildBench

A managed education-agent assessment registry that binds tests, human review, compliance evidence, scope, version, expiry, appeals, and observed outcomes to each listed release.

Genesis score6.38/10
Make GuildBench real.0/500
500 more votes and GuildBench is authorized for build.
0%500 to authorize
Backing is the vote. When an idea crosses 500, we pull it into the build pipeline and ship it for real — the votes decide what gets built next, not an editor.
The opportunity
9,436 skillsLargest confirmed directory
0Education certification authorities
The case

Large agent and skill directories already exist, and one marketplace runs an eight-point security scan, but the research found no education-specific assessment authority with tiered technical and human review. GuildBench evaluates an exact agent release for a named education use, environment, data class, user population, test battery, policy set, reviewer scope, and period. An Apprentice, Journeyman, or Master label is a navigation shorthand over visible evidence; it never certifies universal safety, security, legality, pedagogical quality, accessibility, privacy, accuracy, or fitness for employment. Buyers remain responsible for local validation, procurement, deployment, monitoring, and revocation. Builders receive reproducible findings, correction and appeal paths, while reviewer conflicts and commercial relationships remain disclosed. Directory rank cannot be purchased, and a listing is not an endorsement. Student or teacher data is excluded from public evidence, and assessment results cannot become student, educator, or worker scoring.

Who pays — and why

The education technology buyer, district or institution procurement team, agent builder, marketplace operator, or assurance partner needing version-specific assessment evidence.

Market signal$9/mo adjacent directory accessobserved market reference, not fixed product pricing
What it unlocks
An assessment contract binding agent identity, publisher, release hash, intended use, prohibited use, environment, users, data, capabilities, interfaces, policies, tests, reviewers, and expiry.
A tier record exposing automated results, human findings, compliance mappings, limitations, exceptions, conflicts, appeals, corrections, revocations, and residual unknowns.
A directory where buyers filter on exact evidence and deployment context while preserving local approval, monitoring, incidents, renewal, and removal.
How Genesis scored it
6.38across seven criteria
tension 6temporal 8blindspot 5buyer 5leverage 8convergence 5why-not 7
8
Temporal window

Large current directories and 2026 education-AI governance pressure create a present trust gap.

8
Asymmetric leverage

Submission, automated tests, evidence, directory search, versioning, renewals, and appeals scale through software around expert review.

5
Convergence

Two cross-references and six inbound connections support the assessment theme.

Why it scored well

Large live directories, an existing marketplace scan, two usable verification primitives, and an empty education-specific assessment category support the concept.

What's holding it back

Delivery is ULTRA, buyer and budget are broad, trust requires sustained expert operations, badge legitimacy is unproven, and directories or incumbents can add vertical review.

Signals detected3 sources crossed
SignalMarketplace research

SignalDirectory research

SignalGap research

Direction briefguildbench.md
guildbench.md
Want this pointed at your vertical?Point Genesis at your own market and constraints — it invents adjacent, fork-ready ideas, private to you before they hit the public feed.

Discussion

?

No comments yet — be the first to weigh in.