Verify a decision
Every moderation decision on AVOID.NET is anchored to the Solana blockchain. You don't have to trust us — you can verify cryptographically that we committed to a verdict at a specific moment and have not rewritten it.
How verification works
- We commit. When a moderator accepts/rejects a submission, we serialize the decision into deterministic UTF-8 bytes (
payload_canonical_string), hash it with SHA-256, encode the digest as base58, and write it to Solana inside an SPL Memo v2 transaction. - We store the bytes. The exact bytes we hashed are stored alongside the decision in our database. Anyone can read them and recompute the hash in any language.
- You compare three values. Database hash, your independently-recomputed hash, and the hash inside the on-chain memo. If all three match, the decision is authentic and timestamped.
The on-chain memo format is
AVOID.NET|v1|h:<b58-sha256>|d:<id>|t:<iso>Find a signature on any investigation page's decision log, or run python -m src.verify_decision --signature <sig> for a CLI check.
Decision
review · debank.auction
- Sequence
- #4
- Score
- 0 → 0 (0)
- Cluster
- mainnet-beta
- Slot
- 457569203
- Off-chain at
- 2026-10-16T08:22:57.435Z
- Anchored at
- 2026-10-16T08:23:08.414Z
- Block time
- —
Independent verification
- 1. Database (off-chain)
- 3LTGmy677ZCQaKFjesDnyosc8MB1Pp8de6JH3aeSQEWE
- 2. Recomputed (your browser)
- computing…
- 3. On-chain (Solana memo)
- fetching…
Canonical bytes hashed (1839 chars)
{"actor":"reviewer","artifact_identity":"6da84851b1cff207d4454a341f38d4e2","decided_at":"2026-10-16T08:22:57.300Z","decision":"review","disputed_pct":0.25,"findings_hash":"ExPzJh7b6Br3KY8v9PhEbKq7P3fuKJWNAA1WKzhgAyHB","investigation_id":"027efe27-8bb6-4acc-8596-12b3ac306d6a","mode":"findings_only","new_score":0,"page_content_hash":"50d4f60ec02d2d68e683d6daaf3f25a9a4a7b765a3656d2dfa2e2a900908bba2","page_slug":"debank-auction","prev_score":0,"reason":"The page's central, novel claims — that debank[.]auction is a Zscaler-documented indirect prompt injection (IPI) campaign, that it was tested against 26 LLMs with GPT-5.4 and Claude Sonnet 4.5 misclassifying it absent a trusted reference, and that a companion requests-secure-v2 campaign induced four of 26 models to pay a hardcoded wallet $3 (~0.0012 ETH) — all check out against a direct fetch of the primary Zscaler ThreatLabz post and independent press coverage (SecurityWeek, Infosecurity Magazine). The weakest parts of the page are an apparently fabricated 'updated August 5, 2026' detail (likely a Wayback archive-capture timestamp mistaken for an editorial revision date, appearing in both a section and a timeline entry) and the 'Broader Phishing Ecosystem' section, where two claims are over-attributed to sources that do not actually say what the page says they say (PhishDestroy does not classify the cited domain as phishing/drainer in its own text, and the cited Group-IB research never mentions DeBank or the enumeration mechanism described). The Entity Overview section is independently true but is backed by a source that doesn't actually contain the specifics cited. No accepted community submissions were present to review.","score_delta":0,"sequence_num":4,"submission_content_hash":null,"submission_id":null,"submission_kind":null,"submission_valence":null,"v":1}