Loading ToolBench
MCP server for the StackScan Tech Lookup API - look up the technology stack and the company behind any domain.
StackScan MCP has 6 well-named tools with clear, action-oriented names (lookup_*, check_*). Descriptions are present and moderately detailed (avg ~120 chars), explaining what each tool does and credit costs. However, input schemas are minimal: most tools accept only 1-2 simple string/array parameters with basic type info but no enums, ranges, or validation constraints. Output schemas are entirely undocumented, LLMs cannot predict what fields to expect from responses. Error handling is present in code (ApiResult type, noData flag) but not surfaced in tool definitions. Parameter descriptions are adequate but lack format/constraint details. No tool annotations visible in definitions despite toolAnnotations=true in features.
Check the StackScan credit balance. Free - does not consume a credit. Call this before a batch of lookups so you know how many you can afford.
Batch lookup of companies behind multiple domains. Returns firmographics for each domain. Costs 1 credit per successful lookup (misses are free).
Look up the company behind a domain. Returns firmographics: company name, industry, location, headcount, funding, and more. Costs 1 credit per successful lookup.
Look up the technology stack running on a domain. Returns detected technologies, versions, and confidence scores. Costs 1 credit per successful lookup.
Batch lookup of technology stacks for multiple domains. Returns detected technologies for each domain. Costs 1 credit per successful lookup (misses are free).
Output schemas completely undocumented. LLMs cannot predict response structure (fields, types, nesting). Responses likely include raw API payloads with irrelevant metadata, wasting tokens and risking context window exhaustion.
Input parameters lack validation constraints. 'domain' and 'technology' accept free-form strings with no format, length, or pattern guidance. No enums, ranges, or examples. LLMs may pass invalid values (e.g. 'example' instead of 'example.com').
Error handling not surfaced in tool definitions. Code distinguishes 'noData' (miss, free) from real failures, but LLMs see only success/failure. No guidance on retry strategy, credit implications, or recovery steps in tool descriptions.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 64 | 2026-07-28+ | v2 |
Look up which domains run a given technology. Costs 1 credit per successful lookup.
Tool annotations declared in features but not visible in tool definitions. readOnlyHint, idempotentHint should be explicit in schema to guide agent planning and retry logic.
Batch tools (lookup_companies, lookup_domains_technologies) lack pagination/limit parameters. No guidance on max array size, result limits, or handling of partial failures. Large batches risk context window exhaustion.