MachineReach
MachineReach is a free SaaS tool that scans a public website to assess how discoverable, understandable, and usable it is to AI agents. It evaluates the machine-readable surface of a hostname, including crawl policies, sitemaps, OpenAPI definitions, A2A Agent Cards, MCP evidence, ARD draft linkage, and well-known resources. The scan uses deterministic validation, profile-aware scoring, and bounded semantic evaluation to produce an Agent Visibility Score. MachineReach is a product of PRISMGRID TECHNOLOGIES PRIVATE LIMITED. It checks six areas: discovery paths, machine metadata, capability clarity, interoperability, trust signals, and AI understanding. The scoring model v1.0 is calibrated on more than 20 live hosts; a typical website scores near 40, one validated machine interface around 70–75, and agent-native implementations 90–100. Crawls that are blocked or empty are labeled 'Insufficient evidence' and excluded from ranking calculations. The free tier allows scanning any public hostname without signup and returns a public report with a basic score explanation. Paid access adds an exact remediation plan, hostname history, a verified badge, continuous monitoring, and regression/recovery alerts. Billing is not self-serve yet; users can request paid access by email. The product is explicitly positioned as a way to scan, fix one machine contract, verify, share, and monitor AI readiness over time, not as an uptime monitor.
- Category
- Analytics
- Pricing
- Free
Problem solved
MachineReach solves the problem of websites being invisible or poorly understood by AI agents. It scans public hostnames for machine-readable capabilities and provides an actionable Agent Visibility Score, remediation plan, and monitoring so businesses can make their online presence discoverable and useful to AI systems.
Use cases
- Audit a website's current AI agent discoverability and visibility score
- Identify missing machine-readable metadata such as OpenAPI, A2A Agent Cards, and MCP endpoints
- Verify alignment with emerging standards like the ARD 0.9 draft
- Prioritize fixes for improving AI access through sitemaps, robots, and well-known resources
- Monitor a verified hostname for regressions in agent-facing metadata
- Provide objective evidence of AI readiness to stakeholders or partners
Audience
- Developers
- Product Managers
- Business Users
- Enterprise
- Startups
Pros
- No signup required for public scans
- Free tier available for one-off scans
- Passive GET/HEAD requests reduce impact
- Covers emerging standards including ARD, A2A, and MCP
- Produces a public, evidence-based visibility report
Cons
- Paid access is not self-serve and must be requested by email
- Only public URLs are supported; no private or internal site scanning
- Blocked crawlers yield insufficient evidence
- Score model v1.0 is early and calibrated on a limited set of hosts
- The evaluation methodology is not fully disclosed
Official website: https://machinereach.com/
Canonical LambdaBase page: https://www.lambdabase.com/tools/machinereach