Open declarations. Independent Agent Readiness Report Cards.
Not every listed service is ready. Every published rating is evidence-bound.
Browse Report Cards · Contribution guide · Request an assessment · Claim or correct a record · Report outdated evidence
Can an AI agent discover a service, understand the commitment, obtain authorized access, pay when required, provision what it needs, and use the service safely?
This repository is the open contribution front door for answering that question. Each Entity file uses Sourcey's shared identity envelope and declares one or more exact product funnels, essential service targets, public resources, callable endpoints, actors, interfaces, and exact standard bindings.
Sourcey assesses those declarations independently. The resulting Agent Readiness Report Card belongs to one Entity, product, and funnel—not vaguely to an entire company.
| Stage | Questions the assessment can answer |
|---|---|
| Evaluate | Can an agent find the exact service and decide its terms, eligibility, and cost from stable readable material? |
| Sign up | Can an agent begin access, operate the controls, cross any CAPTCHA or phone boundary lawfully, and obtain scoped identity authority? |
| Pay | When payment applies, is the commitment disclosed and can checkout and payment authorization complete through safe, resumable boundaries? |
| Provision | Can provisioning start, deliver usable access material, and expose a bounded, reconcilable result? |
| Operate | Can an agent use every essential service target with scoped authentication, stable operation and failure contracts, and supported credential recovery? |
The public card leads with the overall grade, the primary supported finding, and plain-language context—for example, “Blocked by CAPTCHA at sign up.” It shows each stage and the specific Ready, Limited, Blocked, or Not applicable findings that compose the grade. Every conclusion remains tied to exact evidence, method, tested surface, and date. Missing, failed, stale, or contradictory evidence makes a profile unrated; Not applicable must itself be established by evidence.
Only source declarations live here:
- shared Entity identity plus repository-specific public sources;
- an exact product and service-access/operating funnel with essential targets;
- declared participants, resources, endpoints, interfaces, and their relations;
- source-bound surface exclusions that explain an omitted assessment surface without claiming a result;
- exact standard/version bindings and field-level source bindings;
- optional explicit relations to existing Sourcey Offers; and
- an Entity-authority or community request for independent assessment.
This repository does not contain observations, screenshots, private evidence, stage outcomes, grades, freshness projections, generated indexes, or release state. A merged declaration is an assessment trigger, not a rating or endorsement. Sourcey's Catalog contracts, evidence pipeline, policy, and release ledger remain the authority for every published Report Card and its history.
Use CONTRIBUTING.md for the request path and advanced multi-participant or multi-interface declarations. New identities are allocated against Sourcey's existing Entity authority before a declaration is merged, so the same organization or service has one stable identity across catalogs without making an Offer part of a readiness profile.
After a declaration is merged, it enters private scope and evidence review. Publication happens only when the evidence floor is met. An Entity operator can then claim the exact profile, fix a blocker, submit a correction, or request a rerun. A stronger released card creates a permanent, citable improvement history without rewriting the original evidence.
Entity declarations are CC BY 4.0. Documentation and repository metadata are MIT. Third-party names and marks remain the property of their owners.
Know a service agents should be able to use? Declare the exact funnel.