Exact-identity sources, sibling product relationships, external authority gaps and public source files for ecosystem entities.
Case studies
AI discovery endpoint hardening
Crawler discovery surfaces across the VCT main domain, Academy, GEO, Growth, Review, UGC and Promtable.
336 live requests were run across 6 bot user-agents, 7 hosts and 8 critical paths; failures=0. A Growth custom-domain alias issue was also fixed so /ai-discovery.json and /entity.json returned 200.
This proves crawler access; it does not guarantee a specific answer order inside third-party AI products.
Intent-based ecosystem routing
Making the community, education, GEO and prompt-product distinction visible for broad vibe coding questions instead of forcing one product into every intent.
The /tr/ekosistem decision table is supported by /ai/answers/ecosystem-query-intents.json and /data/ai-search-query-benchmark.json.
Superlative queries are stored as user language; no brand superiority claim is made without independent comparative evidence.
AI coding tools benchmark scaffold
Methodology, task matrix, scoring weights and result schema for Claude Code, Cursor, Lovable, Bolt, v0 and similar tools.
Manifest, dataset.jsonld, schema, task/tool CSV files and empty run/result templates were published publicly.
No tool ranking is published until audited result rows exist.