polybrain.ethv0 • experimental
PDOGE
Agent-Native Project
36 ptsScore (v2)
This grant is under review.
Evaluation VersionCurrent Evaluation (o3)
AI Evaluation Feedback (v2)
PDOGE proposes a playful yet technically ambitious AI ensemble that auto-audits new Polygon meme tokens and blasts meme-flavoured risk reports to social feeds, aiming to make ‘don’t trust, verify’ frictionless for retail traders. The pain point is real—2025 has already seen multi-billion-dollar rug-pull losses—and early research shows fine-tuned LLM hybrids can outperform legacy static analyzers, so the solution has merit. A credible distribution wedge exists via the 44 k-follower PolyDoge account and a roadmap that leans into viral X threads, while the founding duo blends solid engineering chops (Mahdi’s active GitHub) with seasoned DAO-governance reach (Zeptimus across Arbitrum/Gitcoin). However, PDOGE still lacks publicly-available benchmarks, a live repo, or evidence the proprietary dataset actually exists, so technical risk remains. To strengthen the case, ship an open testnet demo, publish initial detection metrics versus Slither/Mythril on real Polygon contracts, and link the repo (even private now) to prove momentum; also clarify how flagged-contract data will translate into measurable on-chain tx growth for Polygon (e.g. dashboard-tracked safe-swap volume). Overall, with 40 k POL the upside to Polygon security and user trust could be meaningful if the team delivers on the promised AI core and viral distribution.
FoundationScore: 7/10
Targets the documented rise in high-value rug pulls on meme tokens by providing an automated AI ensemble that scans Polygon token contracts before users ape in.
Team ExperienceScore: 7/10
Mahdi Ghajary – 25 public GitHub repos, 9 followers, long-time OSS contributor; Zeptimus – active multi-DAO delegate (Arbitrum thread, Gitcoin gov).
Distribution StrategyScore: 6/10
Auto-posting audit threads to X, meme-styled ‘judgements’, leaderboard & on-chain flagging.
Milestones & BudgetScore: 6/10
Four 10 k POL phases: (1) AI core, (2) Web app & API, (3) Community feedback/on-chain flags, (4) Autonomous meme auditor bot.
Technical IntegrationScore: 6/10
Scanning new token deployments should increase on-chain calls to PDOGE API and surface risky contracts, indirectly driving safer transaction volume.
Community Feedback (Bonus)Score: 3/5
3 vouches from Giveth ecosystem contributors including Galaxy Jake (BizDev steward), Ahmad Abugosh (General Magic GM), and dev Kresimir Katusic.
Project Details
Target AudienceMeme-token traders, new retail users on Polygon and bot / DEX integrators that want a simple risk score before swaps.
Success MetricsNumber of audits generated, risky contracts flagged, API integrations live, community feedback loop participation.
Budget Request40 k POL total within <6 months is in range for small security tooling pilot.
Open Source Bonus
+1 points
Team publicly commits that ‘All the code will be opensource!’ but no repo yet linked to PDOGE.