{"id":"salespeak-ai-buyer-eval-skill","name":"buyer-eval-skill","af_score":33.2,"security_score":41.0,"reliability_score":28.8,"what_it_does":"A free open-source Claude skill that evaluates B2B software vendors by prompting for buyer context, conducting domain-expert question exchanges, optionally engaging vendor AI agents via the Salespeak Frontdoor API, cross-referencing claims against independent sources, and producing a comparative, evidence-tagged scorecard plus risk analysis and demo-prep questions.","best_when":"You are evaluating multiple vendors in a B2B software category and want faster, evidence-tagged comparative analysis that distinguishes vendor-verified vs independent claims.","avoid_when":"You need a strict, documented machine interface (OpenAPI/SDK) for deep automation, or you require guaranteed availability and well-defined reliability/SLA for automated runs.","last_evaluated":"2026-03-30T15:24:58.772258+00:00","has_mcp":false,"has_api":false,"auth_methods":["Salespeak Frontdoor API (used internally by the skill for POST requests when available)"],"has_free_tier":true,"known_gotchas":["Upstream vendor-agent availability may vary (skill behavior differs when a vendor has a Salespeak Company Agent vs not).","End-user auth/controls for upstream calls are not described; agent outputs may depend on upstream response quality and coverage.","No documented interface contracts (tool schemas, retries, error codes) are provided in the README, which can complicate robust automation."],"error_quality":0.0}