In my last post I ran the git logs of Hive's three big frontends through the same audit and let you judge who earns their funding. .This one is different. This is the story of the largest development payout the DHF ever made to a project — told entirely in timestamps: commit dates, hourly payment records, and the team's own posts. I'll warn you now that it doesn't have a villain. It has something more uncomfortable.
Same rules as last time: every date below comes from the chain, the repositories, or the team's own words. Both readings of the evidence are given. Check everything. Point to note that I have stayed in my own little bubble on Hive for a number of years and not being on Discord all that much to see the arguments and who is friends with who. I do not know any of these platform devs personally which means I can be more objective in my approach when it comes to the reviews.
The headline numbers
VSC / Magi Network received 722,084.514 HBD from the DHF — 20,572 hourly payments to @vsc.network, the first on November 1, 2023, the last on May 7, 2026. Three funded proposals: #279 at ~451 HBD/day, #303 at 665 HBD/day (jointly with V4V.app, whose verifiable share — 54,225 HBD in near-daily transfers — was passed on cleanly), and #342 at 1,151 HBD/day for its final year — the largest project allocation on Hive, nearly double Keychain's.
Across the same lifetime: ~7,400 commits over 30 public repositories, 21 contributors, a peak team of about seven, a testnet, a mainnet, a Bitcoin integration, and a rebrand.
And in August 2026: two commits. Total. Org-wide.
The timeline
2021–2023: @vaultec starts alone in 2021; a team forms through 2023. DHF funding begins November 1, 2023.
April 2024: Testnet launches — "VSC only started out with myself, and since then has grown to a team of five." (The launch post also says, in a line that aged strangely: "There will be no mainnet; the testnet will slowly be upgraded.")
2024: Funding steps to 665/day. In September the team begins rewriting the entire node from TypeScript to Go — a major mid-funding re-architecture.
March 31, 2025: Mainnet launches, listed as a market event on CoinMarketCal. Funding steps to 1,151/day in May. The funded steady-state follows: 150–220 commits a month for most of 2025, the strongest sustained output of the project's life.
November 2025 — three weeks that contain the whole story. November 6: rebrand to Magi Network. November 11: Bitcoin mainnet integration goes live, with BTC-HBD and HBD-HIVE liquidity pools seeded with $100–200K of HBD and BTC and the public invited to deposit — "set it and forget it." November 28: the team publishes a remarkable essay, "On Repeated Development Work, Misaligned Incentives, and the Need for Strategic Coordination on Hive." Hold that thought — we'll come back to it, because it's the hinge of this entire post.
And in the same three weeks, the git log records what none of those posts mention: the founder's commits end. Vaultec — 604 commits across four years — trails off in November and December, makes one final commit in January 2026, and is gone. His direct payments from the project account had already stopped in May 2024 — meaning, as far as the chain shows, he kept building Hive's flagship L2 for eighteen months after his last recorded payment. He has never posted about leaving. I'm not going to speculate about why on his behalf; it's his story to tell if he wants to, and the only fair thing to print is the observable record and one respectful open question.
March–June 2026 — the sprint. With the 1,151/day line approaching expiry, output explodes: 625, 712, 1,185, 746 commits a month — a tenfold jump from the winter. Nine new repos in four months. The work is real and hard: BTC vault key rotation, proof-of-authority consensus, ZK header verification, DEX invariant guards, test suites literally named "a compromised operator cannot steal."
May 2026: The old proposal pays its final HBD on May 7. Five days earlier, the new ask posted: proposal #378, 900 HBD/day — a 22% cut — with a pledge of revenue-based reduction "towards $0 USD." Structurally, the most accountability-forward large proposal Hive had ever seen: a working fee engine (8bps protocol fees plus slip-based fees, live since November) explicitly promising to defund itself as revenue grew.
It never crossed the return threshold.
July 2026 — the unravelling, developer by developer. techcoderx (687 lifetime commits): last activity June. Milo Ridenour, the year's workhorse (1,476 commits): from ~190 a month to 21. tibfox (1,270 commits): from 263 in May to 27. What remains in July is @lordbutterfly — 51 commits, his biggest month ever — one other contributor, and a brand-new commit author appearing only that month, named "clauderfly." The last operator, and by every appearance his AI assistant, holding the fort together.
July 23, 2026: the final substantive commit: fix(poa): close RG-1c. A security-audit finding, closed mid-campaign. The three weeks before it read: "eight defects found by this pass," "the halt never armed," fix after fix on the machinery that guards the vaults. Then it stops.
Where the money went
Since "where did the money go" is the question everyone asks first: the chain answers it, and the answer is boring in the best way. Roughly 3.76M HIVE flowed out through exchange deposits — which is what paying salaries looks like, since developers can't eat HBD. Contributors were paid directly on-chain (vaultec ~30K HBD through May 2024; later-era payments to lordbutterfly, disregardfiat, tibfox, and a dozen others). V4V's joint share was passed through daily for two years without a gap. 332K HIVE was powered up to the project account. HiveFest got sponsored. Nodes got paid. Some 127K HBD was converted via @keychain.swap. Over 31 months, 722K HBD averages about 23K HBD a month for a five-to-seven-person team building an L2 with Bitcoin custody — below Western market rate for that engineering. By Hive's standards the number is enormous; by the software industry's it is ordinary. Both of those facts are true and both belong in this post. I found no evidence of misappropriation, and I looked.
The essay that haunts this story
Now back to November 28. The "Misaligned Incentives" post is the sharpest critique of the DHF anyone has published, and it came from inside the house. Read these lines knowing what happened next:
Too many projects "begin not with a market need, a sustainable revenue plan, or cross-project consultation, but with one question: What can I propose to the DHF?" The incentives produce "self-justified, low-impact projects designed primarily to secure funding." Funding flows to "initiatives that show limited usage or no real world usage at all." And: "Double spending isn't only a blockchain threat, it's what we do to the DHF every time we pay for duplicate work because no one coordinates."
Every word of that diagnosis is defensible — my last post's data on three teams building three codebases for one social layer is evidence for it. And yet:
They described their own chart. "Value creation becomes secondary to merely qualifying for DHF funding," wrote the project whose commit curve is the purest proposal-shaped mountain I have found on this chain — output that multiplied into every ask and collapsed when the last one failed.
Their own criterion killed them. They demanded funding go only to "outcomes with measurable impact" and away from projects with limited real-world usage. Five months later, stakeholders applied precisely that test to a 900/day ask for a bridge with thin usage — and declined. The community took the prescription. Magi was the patient.
And the self-serving reading must be given its due. A call to "consolidate talent" under "a coordinated framework," published by the chain's largest funded project mid-rebrand, also reads as: fund fewer things — fund us. Their own Go rewrite of their own TypeScript node is a year of repeated work by their own definition. None of this makes the diagnosis wrong. It makes the author compromised, which is the most human thing in this whole story.
The two questions that are still live
This is the part that matters more than any of the history, and I'm addressing it to @lordbutterfly and the remaining team directly — as questions, because only you can answer them.
1. What is in the vaults right now? The November 11 launch seeded $100–200K into BTC-HBD and HBD-HIVE pools and invited public deposits into what was marketed as non-custodial with "no third-party risk." The July commit log shows those funds are held by threshold-signature vaults whose security hardening — the operator-theft protections, the exit-halts — stopped mid-audit on July 23. "Trust-minimized" was always the honest term; the risk is real, distributed across operators, and secured by code whose audit is unfinished. How much user money is currently on those vaults, who operates them today, and is there a wind-down or withdrawal plan? This is a risk-disclosure question, not an accusation, and a clear answer would serve everyone — including you.
2. What happens to the token promise? Early liquidity providers were told they'd be rewarded at the MAGI token launch, "scheduled some time next year, 2026." It's August 2026 and development has effectively stopped. People who parked BTC partly on the strength of that pledge are owed a status, whatever the status is.
For the record: I've mirrored all 30 public repositories in full — every branch, every commit, cryptographically hashed — as of August 10. Whatever happens to the org, the record of what 722K HBD built is preserved and verifiable.
Two readings — keep both
The failure reading: Hive's most expensive project produced a chain that froze three commits short of its security audit; the moment the payments stopped, so did the work; and the output curve tracked the proposal calendar more purely than any project I've examined. If you want the DHF's pathologies in one dataset, this is it.
The other reading: this is simply what buying software looks like. Paid professionals built at pace for two and a half years, shipped a testnet, a mainnet, and a Bitcoin integration, were offered up at a 22% discount with a self-defunding revenue model — the only large proposal on the chain with a working business engine attached — and the community said no. The developers then did what workers everywhere do when the pay stops: they left. Declining the renewal was a defensible choice under the project's own published standard. The result of that defensible choice is a half-finished bridge with money on it and one man and his AI closing audit findings until July 23.
Both readings are true. What they share is the actual lesson, and it isn't about this team at all: the DHF has an on switch and an off switch and nothing in between. No milestone escrow. No completion bond. No wind-down funding for a project custodying user assets. No mechanism to land a 722,084-HBD investment softly instead of dropping it. Magi correctly diagnosed the fund's inability to coordinate what it starts — and then became the first major demonstration of its inability to finish what it funds.
We should fix that before we fund the next bridge. Because the next one will end the same way, and we'll have paid for the lesson twice — and as a wise post once put it, double spending isn't only a blockchain threat.
Verify me
The funding figures come from proposal_pay virtual operations in @vsc.network's account history — 20,572 of them, cross-checked against the known daily rates to within rounding. The commit data comes from full mirrors of every public vsc-eco repository. The quotes come from the team's own posts, all still on-chain. Run the numbers yourself; if I've erred, say so in the comments and I'll correct it prominently, same as always.
Analysis as of August 12, 2026. Funding data: on-chain proposal_pay operations and account transfers. Commit data: all branches of all 30 public vsc-eco repositories, bots excluded. Nothing in this post should be read as an allegation of misconduct against any individual; it is an account of what a public treasury's records show, published so the community can decide what to change.
Great breakdown. I was never super up to date with the project but was aware of it's existence and felt it could be an X Factor for this ecosystem.
Is the service usable or is the liquidity so low that it can't really be used at scale?
Service is usable . At scale I am not sure. It comes down to pool depth which I do not think they have yet.
The liquidity is low and we are not currently market viable since the protocol is not finished. It was explained in the proposal at what stage we are.
Ok cool. Hopefully there will be a path to getting it operational at it's full capacity. It could open up all kinds of posibilities.
You will see some really cool possibilities soon. 🙂
Awesome 👍
THis is sad because there was really a lot of hope this was going to fix things. I think even Dan was behind it.
Tough when crypto use is declining at a rapid pace. Platforms like this need traffic and a large userbase. Hive doesn't have that.
I'm not sure it ever will. If you look at other projects they are going through much of the same thing. Use case means a whole lot of nothing when people are just playing the markets.
It got it down to 2000 words Bozzy 😃
Oh, don't worry, I noticed! :)
I would really love to see this project succeed on some level.
Yeah i think we all
Would like to see it do well @daveks. Would be good for the place
Would like to have others chime in to this post, ie: vaultec, blocktrades, tibfox..
Over the last couple of weeks I’m starting to really see blocktrades logic on what he supports and what he doesn’t support.
Tell me more about that?
So, magi isn't gonna save us after all?
Doesn't look like it now anyway. These crypto bridge platforms be compli-ma-cated.
Good luck! Hopefully u get some answers.
Reblogged, for awareness.
Thanks Qasim.
cheers guys.
Inspector Blanchy on the beat again. This scooby-doo dhf caper will be unmasked.
Not too bad for an AI driven analysis. Usually it does halucinate or repeats itself from different angles. Was surprised with what it got right. A few things it got wrong or didnt have enough context but I think its a fair assesment.
Just a few notes, nothing major:
These 2 last commits you saw are major development additions worked on for weeks.
Appreciated and noted on the pool launch date and the seeding figure. Those came from your own announcement post, which I read as a launch rather than a plan; I'll correct that, and it's a fair catch.
On the rest, a few of these are explanations rather than corrections, so I'll take them as they are bullet pointed:
The 10x output — I gave both readings in the post: your account (changed approach, AI-assisted dev, a push to revenue) alongside the observation that the surge tracked the funding cycle. Both can be true. Your explanation is in there; readers can weigh it against the dates.
Commits vs LOC — agreed, and I said the same throughout: commit counts mislead, which is exactly why the post leans on the payment records, not the commit graph, for anything load-bearing. Those payment dates aren't a proxy for anything — 722,084 HBD to @vsc.network, first payment Nov 2023, last May 7 2026. That part isn't in dispute.
The private repo is the one point I can't do anything with. If significant work isn't public, nobody outside the team can verify it — and a publicly-funded project keeping its development private is the transparency question the whole post is about, not an answer to it. I can only audit what's visible, and I was clear about that limit.
For what it's worth, I checked the bridge before replying: it's live, gateway's processing, withdrawals completing as of this week. Credit where it's due — it's still running. Happy to be proven wrong by results; that's usually the part worth waiting for and hopefully will benefit the Hive Community in the future and get us out of this hole we are in .
And finally 90% of all analysis is AI driven these days. The halucinate or repeats can be managed effectively in premium models like Fable by a second pass or a different premium model. The context was pretty solid but always room for improvement. Cheers for the reply.
One more thing, kind of a personal thing. That chart there you put up. That 10x is something I am incredibly proud of. Proud of myself and proud of my team for what we acheived. If you read the proposal post again and put the dates together, that is exactly when I took over Magi. Exactly on that red line.
The reason I have a problem with whats written is the missatributed motivation presented in the post. It was not about "chasing DHF funding" via forever proposals that never end and never change anything, never finish anything, never deliver anything. Just add more submenus and login options.
That is the exact pathology on Hive I was trying to end. I was chasing self-sustainability.
I meant just in general for this type of work. I did this exact same analysis for all proposals a few weeks back. My findings were a bit different. I did look at some other stuff your AI probably missed. hehe.
Yeah, sure. Just saying as a response that stuff is being worked on.
I meant just in general for this type of work. I did this exact same analysis for all proposals a few weeks back. My findings were a bit different. I did look at some other stuff your AI probably missed. hehe.
Any plans to resubmit the proposal at a lower ask?
Probably not. We have already been building this on a budget thats 95% smaller then the size of budgets those with our similar tech work with.
And we dont have the luxury of arbitrarily changing proposal amounts, we are actively building. Magi is not like the "forever proposals" which are essentially in perpetual maintenance so they can take whatever DHF might give them.
We cant. We arent completed yet.
We cut our ask by 22%, which cut marketing, and proposed a plan towards self sufficiency that I think was very solid. The community decides.
Thanks for the reply. Looking forward to the success of magi, and I like the rebrand btw.