Reading this thread, what stands out is that nothing here looks improper. The RFP is public, the deliverables exist and are open source, quarterly reports were published, and Albert answered within hours with a coherent explanation for the early payout.
And yet three posts in, the community still can’t independently verify whether 160K DOT bought what it was supposed to buy… because the original scope, milestone definitions and acceptance criteria were never on-chain in the first place. Everything we’re relying on lives in a website, a quarterly PDF and the memory of the curators who made the call.
That’s the real finding. This is roughly the good case: competent curators, real deliverables, a responsive answer. If verification is still this hard in the good case, we have no mechanism at all for the bad one, we just have trust in the individuals holding the curator seats, which happened to be well placed this time.
Three structural gaps are visible from this thread alone. Scope and acceptance criteria were never committed on-chain before funds moved, so there’s no reference to audit against. A tranche was released for work that was still in progress, on a judgement call made under closure pressure, with no delay window in which anyone could have raised an objection. And whatever the outcome turns out to be, it carries no consequence for the curators either way… approval and outcome are financially disconnected.
Worth saying that “remove the bounties pallet” doesn’t fix any of those three. Routing this through a direct referendum instead would have produced the same missing acceptance criteria and the same unverifiable outcome, just spread across a much larger group with no one specifically answerable. The problem isn’t the pallet, it’s that delegated spending has no accountability layer.
I’ve been building one (why? because i can). Five FRAME pallets are live on Paseo… on-chain milestone definitions before funds move, escrow that withholds part of each tranche until final delivery, a mandatory delay between curator quorum and any large release, and curator reputation scored retrospectively when a project reaches its terminal state rather than at approval. It’s the same failure set this thread surfaced, addressed structurally. Full write-up here: Bounty 22 was close to the good case. We still can't verify it
Disclosure: this is my own project (Two Pebbles Ventures). Happy to have the design attacked… that’s more useful to me than agreement.