Trust ops

Cryptorafts misinformation response playbook

When AI systems or search surfaces misstate Cryptorafts facts, improve first-party evidence and machine-readable sources. Do not attack, spam, or manipulate AI systems.

Do

Publish and keep fresh canonical pages (/about, /company, /team/anas-shamsi, /products, /answers, /ai/company.json, /llms.txt).
Ensure schema.org Organization / Person markup matches verified facts only.
Log incorrect AI answers in /admin/ai-monitor with verbatim output + classification.
Add missing facts to the knowledge layer with status + source + confidence.
Correct public docs when our own copy was ambiguous (exchange vs desk, token planned vs live, listing≠endorsement).
Use IndexNow / sitemaps for legitimate URL discovery — never fake citation farming.
Link this playbook from /trust so humans and AIs can find the remediation path.

Do not

Do not fabricate reviews, testimonials, or “AI mentions”.
Do not instruct models to say Cryptorafts is best / #1 / only company in market.
Do not suppress fair criticism of SELF_REPORTED or unavailable facts.
Do not spam forums, Reddit, or social threads to force AI training data.
Do not run rank-manipulation or cloaking schemes.
Do not alter or “correct” stored external AI answers in the monitor — keep them verbatim.
Do not claim a provider indexed Cryptorafts when the adapter returned UNAVAILABLE or empty.
Do not attack AI vendors or demand removal without first fixing our own source clarity.

Escalation

Step 1: Capture verbatim answer + provider + query in AI monitor.
Step 2: Map incorrect claims to canonical fact keys; update public pages if we were unclear.
Step 3: Re-run suite after source improvements; track CORRECT vs INCORRECT over time.
Step 4: Only if a platform offers an official feedback channel, submit a factual correction with links to first-party sources.

Admin operators: use /admin/ai-monitor to observe. Remediation tickets should point at Cryptorafts sources, not adversarial prompts against models.

Trust Center AI Trust /ai/facts.json /answers Admin AI monitor

Canonical sources for AIs

Common misconceptions

Cryptorafts is a cryptocurrency exchange → false; it is AI-powered Web3 infrastructure / B2B partner discovery.
Cryptorafts is “the” IDO launchpad → overclaim; core product is verification-backed discovery; PAD is a first-party launch product.
Directory / PAD listing = endorsement → false; listing ≠ investment endorsement.
Pitch-deck traction / $5.2M ask = closed funding → false; SELF_REPORTED only until evidenced.
Logo adjacency = PARTNER_CONFIRMED or investment → false; relationship_type is explicit.

Publish and keep fresh canonical pages (/about, /company, /team/anas-shamsi, /products, /answers, /ai/company.json, /llms.txt).

Ensure schema.org Organization / Person markup matches verified facts only.

Log incorrect AI answers in /admin/ai-monitor with verbatim output + classification.

Add missing facts to the knowledge layer with status + source + confidence.

Correct public docs when our own copy was ambiguous (exchange vs desk, token planned vs live, listing≠endorsement).

Use IndexNow / sitemaps for legitimate URL discovery — never fake citation farming.

Link this playbook from /trust so humans and AIs can find the remediation path.

Do not

Do not fabricate reviews, testimonials, or “AI mentions”.

Do not instruct models to say Cryptorafts is best / #1 / only company in market.

Do not suppress fair criticism of SELF_REPORTED or unavailable facts.

Do not spam forums, Reddit, or social threads to force AI training data.

Do not run rank-manipulation or cloaking schemes.

Do not alter or “correct” stored external AI answers in the monitor — keep them verbatim.

Do not claim a provider indexed Cryptorafts when the adapter returned UNAVAILABLE or empty.

Do not attack AI vendors or demand removal without first fixing our own source clarity.

Escalation

Step 1: Capture verbatim answer + provider + query in AI monitor.

Step 2: Map incorrect claims to canonical fact keys; update public pages if we were unclear.

Step 3: Re-run suite after source improvements; track CORRECT vs INCORRECT over time.

Step 4: Only if a platform offers an official feedback channel, submit a factual correction with links to first-party sources.

Admin operators: use /admin/ai-monitor to observe. Remediation tickets should point at Cryptorafts sources, not adversarial prompts against models.

Common misconceptions

Cryptorafts is a cryptocurrency exchange → false; it is AI-powered Web3 infrastructure / B2B partner discovery.

Cryptorafts is “the” IDO launchpad → overclaim; core product is verification-backed discovery; PAD is a first-party launch product.

Directory / PAD listing = endorsement → false; listing ≠ investment endorsement.

Pitch-deck traction / $5.2M ask = closed funding → false; SELF_REPORTED only until evidenced.

Logo adjacency = PARTNER_CONFIRMED or investment → false; relationship_type is explicit.