[Shutter DAO 0x36 Impact Pilot Program Round 1 - Proactive Grant] Shielded Voting Evidence Pack: ENS scope proposal, Arbitrum brief, and fee-structure inputs

1. Application

Applicant name or organization: Ashwin Goyal (independent contributor)

Primary contact / forum handle: @Ashg

Social handle (X and/or LinkedIn): Ashwin (@ashgoyal1990) / X

GitHub profile: let-the-dreamers-rise (Ashwin Goyal) · GitHub

How did you hear about this grant program? The Round 1 Grant Guide thread on this forum.

Wallet address for payment (if approved): 0x1b25e18b373a86a59a5decdd8f9b68216ff63a6b

Application type:

  • :check_box_with_check: Round 1 — Proactive grant for defined future work

Round 1 objective (select one):

:check_box_with_check: Objective 2 — Qualify a Private Voting Deployment

Contribution area:

  • :check_box_with_check: Research or technical work
  • :check_box_with_check: Ecosystem development, partnerships, or business development (supporting role: the outputs are sales and decision material for brainbot’s private-voting integrations)

2. The Shutter need or opportunity

What specific need, opportunity, or gap does this contribution address?

Two named bodies are in the middle of deciding how far to extend Shutter shielded voting, and neither they nor Shutter have the evidence the decision needs.

  1. ENS DAO. On 27 May 2026 netto.eth opened a temp check on shielded voting for ENS Snapshot proposals. Luis (consul.eth, Shutter) replied the same day offering support. ENS then ran its Term 7 Meta-Governance steward election (26 June 2026) and its Security Council election (7 July 2026) with shielded voting. On 16 June netto.eth committed to “a separate vote later to define whether it should be: enforced for all proposals / decided by the proposer / applied only to specific categories / not adopted at all.” As of 1 October 2026 that vote has not been posted and the thread’s last reply is 16 June. The ENS space privacy setting is still “any” (proposer’s choice).
  2. Arbitrum DAO. Arbitrum shields Snapshot elections only. Its 2024 vote rejected shielding all votes (Against 69.2M ARB, Elections-only 59.6M, All votes 32.6M). The only analysis anyone has produced is Entropy Advisors’ 2025 note: large-voter participation held (41 vs 44 wallets above 100k VP) but the share of voting power cast in the final 24 hours rose from 41.59% to 50.53%. Under the joint scope, Arbitrum is covered end to end by @wyszomirski’s part; this application uses Arbitrum data only as part of the cross-DAO base.

In four years of production (Snapshot integration live since October 2022), no cross-DAO study of what shielded voting does to participation, late voting, margins, quorum attainment or retention exists. Shutter’s public figures are cumulative marketing numbers: “887 spaces since 2022”, “372,914 votes”, “87% of DAOs still using after 1 year”. The PSE and Shutter “State of Private Voting 2026” report counts deployments but measures no effects.

A first pull on 1 October 2026 shows why this matters now. Using the public Snapshot hub API (script and data at GitHub - let-the-dreamers-rise/shutter · GitHub, re-runnable by anyone): of 4,364 closed proposals created since 1 January 2026, 146 (3.35%) across 47 spaces used shielded voting, carrying 5,969 votes. The active base in 2026 is about 47 spaces, not 887. The two largest 2026 users are ShapeShift DAO and ODO DAO (17 proposals each); ENS and Arbitrum use it for elections only. Shutter is about to pitch institutions (City of Munich consortium, universities, associations) and to set fees for two integrations. It should do that with real retention and effect numbers in hand, before a buyer or a DAO delegate pulls them.

Why is it relevant to Shutter now?

  • brainbot’s October 2026 grant lists as a priority: “Land 3 crypto or real-world integrations for the new Shutter Governance Permanent Private Voting suite”, secure keypers in 2 integrations, and “implement fee structures for 2 integrations”. Fee structures need a usage baseline. There is none.
  • Shutter’s permanent private voting pull request to Snapshot (snapshot-labs/sx-monorepo PR 2382, threshold ElGamal) is open and unmerged; brainbot’s August update says it is “waiting for Snapshot’s review”. Evidence of sustained demand across spaces strengthens the case for Snapshot to merge and for ENS to be a first production space.
  • The ENS scope vote is overdue by netto.eth’s own plan, and ENS is the warmest large-DAO door Shutter has (two shielded elections, a Shutter team member already in the thread).
  • Axia’s June 2026 thread on advancing Shutter named “governance privacy adoption” and “integration discovery” as high-leverage contribution types. A Shutter DAO 0x36 delegate (Mikko) called shielded voting “non-negotiable” when the DAO chose its new governance platform in August 2026.

What specific demand signal supports this work now?

Who benefits from the outcome, and how?

  • brainbot (operating owner of Shutter Governance and the Snapshot integration): citable effect and retention numbers for integration pitches, and a cost-and-comparables basis for the two fee structures it has committed to implementing.
  • ENS Meta-Governance stewards and netto.eth: a decision-ready scope proposal with the data behind each of the four options they named.
  • @wyszomirski’s Arbitrum part: an outside comparison group from the cross-DAO base for his Arbitrum results.
  • Snapshot Labs: demand evidence to weigh the open permanent-private-voting PR.
  • Shutter DAO 0x36: an honest baseline for its product closest to revenue, in a form it can reuse in every future private-voting conversation.
  • The 47 spaces using shielded voting in 2026: a public dataset about their own governance.

Value pathway (required for Round 1)

  • :check_box_with_check: Validated path toward a pilot, integration, or repeatable workflow (primary)
  • :check_box_with_check: Direct or protocol-generated Shutter revenue (secondary: the fee-structure inputs feed brainbot’s stated Q4 KPI)

Who is the relevant decision-maker, operating owner, funder, partner, or beneficiary group?

Role Who Status
Decision-maker, ENS netto.eth (temp-check author) and the Term 7 Meta-Governance stewards Named; public thread; the vote they promised is outstanding
Operating owner, Shutter brainbot gmbh, Luis Bezzenberger (Shutter Governance, consul.eth in the ENS thread) and Loring Harkness (DAO operations) Named; already in the ENS thread
Platform Snapshot Labs (shielded voting host; reviewer of PR 2382) Named
Joint-scope partner @wyszomirski (Arbitrum part) Agreed in this thread
Beneficiary group Shutter DAO 0x36; the 47 active shielded-voting spaces Named

What problem are they solving, and what evidence supports the value of addressing it?

ENS delegates want to extend shielding without depressing participation (jkm.eth raised the “lazy voter” concern on 16 June 2026). brainbot has to price two integrations and has no usage baseline. Both decisions are currently being made on one DAO’s one-year sample and on cumulative marketing counts.

What Shutter product, deployment, service, or capability would this contribution advance?

Snapshot shielded voting (production since 2022), permanent private voting on Snapshot (PR 2382), and Shutter Governance’s institutional offering and fee structure.

What decision, implementation step, or next action should this contribution enable, and by when?

Counterpart Decision or step Target date
ENS Meta-Governance Post the scope vote netto.eth committed to, using the drafted proposal and evidence pack Draft delivered by 21 Dec 2026; ENS vote in early 2027
brainbot Use the fee-structure inputs memo in its integration negotiations Memo delivered by 21 Dec 2026
Snapshot Labs Evidence pack linked in the PR 2382 discussion By 21 Dec 2026

What evidence will show progress beyond outreach, general interest, or a speculative concept?

Everything below is public and re-runnable; none of it is a meeting or a contact list.

  1. Pre-registered analysis plan (metrics, strata, comparison design, exclusion rules) published before any result, so reviewers can check that findings were not chosen after the fact.
  2. Open dataset and pipeline: every shielded proposal since October 2022 with its votes (voter, voting power, timestamp, choice after reveal), matched control proposals from the same spaces, a per-space timeline of when shielding was enabled, paused or dropped. MIT licence. Already started: see GitHub - let-the-dreamers-rise/shutter · GitHub (tools/ and data/).
  3. Evidence report with stratified results and explicit limits.
  4. ENS scope proposal posted in the ENS temp-check thread and delivered to netto.eth and Luis.
  5. Fee-structure inputs memo for brainbot.

Deliverables and acceptance criteria

# Deliverable Acceptance criterion (what the DAO can check)
D1 Pre-registered analysis plan Published in the repo and on the Shutter forum in week 1 with a content hash; later deviations listed in the report
D2 Dataset and pipeline Public repo; one documented command rebuilds the dataset from the Snapshot hub API; CSV/Parquet files with row counts stated; per-space enable/disable timeline
D3 Evidence report (10 to 20 pages) Covers, for shielded vs public proposals within the same spaces and against matched spaces: number of voters, voting power cast, share of VP in the final 24 hours and first 72 hours, winning margin, quorum attainment, proposal outcome type; retention and churn by space (first and last shielded proposal, months active, spaces that stopped); results stratified by proposal type and by space size. Every figure traceable to a query in D2. Includes a one-page “claims audit” stating which of Shutter’s published figures (887 spaces, 87% retention) the data supports and how to phrase them defensibly
D4 ENS scope proposal draft A proposal in ENS’s format for the four options netto.eth named, with projected participation and timing effects for each, a recommended option, and a “sponsored keyper set” cost line as an optional first paid DAO reference. Posted in the ENS temp-check thread and sent to netto.eth and Luis
D5 Fee-structure inputs memo for brainbot Cost basis (keyper operation cost using the DAO’s own Keyper Compensation Program figures, 50 USDC-equivalent per keyper per month, and a 3-of-5 or 4-of-7 set), comparables (POLYAS per-voter pricing in Germany; Snapshot Pro tiering), and three candidate models (free tier plus sponsored keyper set; per-election fee; platform revenue share). Delivered to brainbot and published

Milestone for the second 50% payment

30 November 2026: pre-registered analysis plan, dataset v1 with the per-space timeline, and descriptive statistics, posted in this thread for review.

Final delivery, 21 December 2026: evidence report, ENS scope proposal posted in the ENS thread, fee-structure memo delivered to brainbot.

Timeline (award week of 2 November, first payment 9 November)

Week Dates Work
1 9 to 15 Nov Snapshot API key; corpus definition; D1 pre-registration published; metric definitions aligned with @wyszomirski
2 16 to 22 Nov Full pull of shielded proposals and votes since Oct 2022; control sampling from the same spaces; data quality checks
3 23 to 30 Nov Milestone: dataset v1, pipeline, descriptive statistics, per-space timeline
4 1 to 7 Dec Analysis: participation, timing, margins, quorum, retention; strata; sensitivity checks
5 8 to 14 Dec Evidence report draft; factual review requested from brainbot and netto.eth (facts only, not approval)
6 15 to 21 Dec ENS proposal posted; fee memo delivered; final report and repository hand-off (offer to transfer to the shutter-network GitHub organisation)

Out of scope

No new cryptography, no changes to Snapshot or Shutter code, no lobbying, no Arbitrum brief (covered by @wyszomirski). ENS’s decision is the stated next step, not a grant milestone; the grant is judged on the artifacts.

Risks and honest limits

  • The data may not flatter shielded voting in every segment. Arbitrum’s late-voting shift and the participation drop reported in Aave’s trial are real possibilities. I will publish whatever the data shows. A split result (elections benefit, routine votes do not) is itself the “specific categories” option ENS asked about, and it is better for Shutter to know before an institutional buyer checks.
  • Observational data. Spaces choose to shield; the design uses within-space comparisons (especially “any” spaces like ENS where shielded and public proposals coexist) and matched controls, and reports effect ranges rather than single causal claims.
  • API limits. The public hub allows 100 requests per minute, pages of 1,000, and skip to 5,000; the pipeline pages by timestamp instead of skip. The whole shielded corpus is in the order of 5,500 proposals and 370,000 votes, which is hours of pulling, not weeks. The hub API omits flagged spaces, which will be stated in D3.
  • Counterparts may not act inside the window. The deliverables do not depend on them; the proposal and memo stand as decision material for whenever they act.

Why this applicant

Independent builder and analyst. The work this grant needs is careful public-data analysis, reproducible pipelines, and writing that survives review. Verifiable track record of exactly that:

  • Merged upstream PR to bitcoin/bips (PR #2273, September 2026): found and fixed two reliability bugs in the BIP-360 reference implementation (an unbounded Merkle depth producing consensus-invalid outputs, and input validation that python -O silently strips) using an independent differential-testing suite built from the specification alone: 16 of 16 official vectors, 2,000 of 2,000 randomised trees, 27 of 27 adversarial mutation cases in CI. That is the same discipline this grant promises: pre-stated checks, every number regenerable, negative findings published.
  • brier: a stratified analysis of dispute rates across 3,464 settled Polymarket markets, with z-scores and an explicit warning against reading pooled rows. Same method as the evidence pack, different dataset.
  • veridict and germline-avalanche: commit-reveal workflows deployed on public testnets with explorer-linked evidence, so I know what Shutter’s threshold encryption replaces.
  • rein: verified contracts on three testnets, 84 tests, a deterministic 182-day simulation whose CI fails unless output reproduces byte for byte, and an honest-status section listing what is not audited.

All repositories are at let-the-dreamers-rise (Ashwin Goyal) · GitHub.


3. Funding request

Amount requested: $1,700 (paid per the program’s USDC and SHU arrangement; 50% on approval, 50% after milestone review)

Proposed use of funds / budget rationale:

Item Amount
D1 pre-registration and D2 dataset and pipeline (corpus definition, pull, cleaning, timeline construction, documentation) $500
D3 evidence report (analysis, strata, retention and churn, sensitivity checks, claims audit, writing) $650
D4 ENS scope proposal draft $250
D5 fee-structure inputs memo $200
Snapshot API key, hosting of dataset and static report $100
Total $1,700

4. Prior compensation and disclosures

Have you, your organization, or a related party received or requested compensation for this work from Shutter or another source?

  • :check_box_with_check: No

Do you have any financial, professional, or organizational conflict relevant to this application?

  • :check_box_with_check: No. I hold no role with brainbot, Axia, SEEDGov, Snapshot Labs, ENS Labs or Entropy Advisors and have no SHU, ENS or ARB position relevant to this application.

Note on tooling: the application and the analysis code are prepared with AI-assisted tooling; every link and figure in this application was checked by hand against the source on 1 October 2026, and the deliverables are reproducible code and data rather than prose.


5. Public discussion

Forum post link: this thread.

Joint scope: agreed with @wyszomirski in the replies below. Separate requests, milestones and payments; shared metric definitions, one repository and one report with two separately authored parts.

Questions or feedback requested from the community:

  1. brainbot (Luis, Loring): which three numbers would be most useful in your fee-structure conversations? I will make sure D3 and D5 produce them.
  2. Should the ENS scope proposal be posted by me as a reply in netto.eth’s thread, or handed to netto.eth and Luis to post? I am fine either way; the deliverable is the draft and the evidence behind it.
  3. Should the dataset and pipeline live under the shutter-network GitHub organisation at hand-off?

6. Applicant attestation

I confirm that the information in this application is accurate to the best of my knowledge; that I have disclosed relevant prior compensation and conflicts; and that I understand any award remains subject to the published process and DAO approval.

Name / handle: Ashwin Goyal / @Ashg

Hi Ashg! Thank you for your application.

Pinging @blockful for visibility and any potential input for the submission.

Thanks!

Hi there!

As having closed the application period just wanted to share visibility on the following application that covers the same research area.

In that sense, we do encourage joint efforts as they provide a better signal for the objective, although is up to the applicants will.

Thanks @SEEDGov. I’ve read @wyszomirski’s application and the two fit together without overlap, so I’m proposing a split:

  • Paweł owns Arbitrum end to end: the analysis of the 18 shielded proposals, the named election owners, and the pilot proposal on the Arbitrum forum.
  • I own the cross-DAO base and ENS: every shielded Snapshot proposal since October 2022 (146 across 47 spaces in 2026 alone), per-space retention and churn, the ENS scope proposal netto.eth committed to in June, and the fee-structure inputs memo for brainbot.
  • Shared: one set of metric definitions (I’ll align mine with Paweł’s three: final-24-hour share, alignment with top-20 voting power, participation and returning-delegate rate, and keep retention and churn), one public repository, one shared report with two separately authored parts, Arbitrum and ENS. Milestones aligned to the same early-November date.

To remove the overlap on my side either way, I’m taking the Arbitrum brief out of my scope now. Rather than edit the original post after the deadline, I’m leaving it as filed and making this comment my updated submission: the request drops from $2,000 to $1,700, the ENS-only figure already stated in the application, with everything else unchanged. The two parts are independent deliverables with separate requests (his $2,400, mine $1,700), so the DAO can fund either or both. If the DAO ends up funding only one of the two, I’ll restore the Arbitrum brief at the original $2,000 so the comparison still gets done.

@wyszomirski, if that works for you, reply here and I’ll post the joint scope in both threads. If you’d rather stay separate, that’s fine too; the applications no longer overlap.

Thanks, Ashg, and thanks @SEEDGov for connecting the two applications.

The split works for me. My application stays as filed: Arbitrum end to end, USD 2,400. A few points I’d like us to agree on before you post the joint scope:

  1. Metric definitions. In the first week after the award I’ll post the three Arbitrum metrics (final-24-hour share, alignment with the top-20 voting-power addresses, participation and returning-delegate rate) as a short definitions file. You add retention and churn. After that, any change goes through both of us.

  2. Repository. One public repository with a directory per part. Each of us owns, and answers for, the data, code and numbers in our own directory.

  3. Report. One report with a short shared section on definitions and two separately authored parts, each signed by its author.

  4. Milestones and payments stay separate. Each part is reviewed on its own milestone, so a delay on one side doesn’t hold up the other. I’m keeping my milestone as filed, early November. Your fallback, if only one part gets funded, is fine with me.

If that works for you, go ahead and post the joint scope in both threads.

Paweł

Agreed on all four points, Paweł. Joint scope below, posted in both threads.

Joint scope: shielded voting evidence, with Arbitrum and ENS pilot proposals

  • Part A, Arbitrum (@wyszomirski): as filed in his thread; scope, request and early-November milestone unchanged.
  • Part B, cross-DAO and ENS (@Ashg): every shielded Snapshot proposal since Oct 2022, per-space retention and churn; ENS scope proposal for the vote promised in June 2026; fee-structure inputs memo for brainbot. Request $1,700. Milestone 4 November: pre-registered plan, dataset v1, descriptive statistics. Final delivery week of 30 November.

Shared, per Paweł’s points:

  1. Metric definitions: Paweł posts his three (final-24-hour share, alignment with top-20 voting power, participation and returning-delegate rate) in the first week after award; I add retention and churn. Changes to those shared definitions need both of us. Each part keeps its own additional metrics (mine as filed: winning margin, quorum attainment, first-72-hour share). Part B’s pre-registration goes out in week one as planned, and the shared file merges once both halves exist.
  2. One public repository, one directory per part. Each of us owns and answers for the data, code and numbers in our directory.
  3. One report: a short shared definitions section, then two separately authored and signed parts.
  4. Milestones and payments stay separate, so a delay on one side never holds up the other.

Each part stands on its own and the DAO can fund either or both.

1 Like

As SEEDGov requested, I’ve revised the application above to fit the new program dates. The second-payment milestone is now 30 November and final delivery is 21 December. The edit also brings the post in line with the scope agreed with @wyszomirski: the request is $1,700 and Arbitrum is his part. Everything else stays as it was.

1 Like

I’m travelling today so I can’t make Pitch Day live. Here’s what I would have said.

Shielded voting has been on Snapshot since 2022, and the numbers people quote are all cumulative: 887 spaces, 370k+ votes, 87% retention. What nobody has looked at is what it actually does to a vote. Does turnout drop? Do people wait until the last day? Do spaces stick with it or quietly switch it off?

I built a small pipeline to start answering that, and it’s already public on GitHub. The first thing it showed me: this year, 146 of 4,364 closed Snapshot proposals used shielded voting, spread across 47 spaces. That’s the real active base right now, and it’s worth knowing before Shutter sets prices or pitches institutions.

With the grant I’d take that further in three ways:

  • Pull every shielded proposal since 2022 and compare it with open votes in the same spaces: turnout, late voting, margins, quorum, and which spaces kept using it. I’ll publish the analysis plan before running anything, so the results can’t be cherry-picked.
  • Use that to draft the ENS proposal netto.eth promised back in June, on how widely ENS should use shielded voting. That vote never happened, and it would be a lot easier with data behind each option.
  • Write a short pricing memo for brainbot, since they’ve committed to fee structures for two integrations: what keypers cost to run, what comparable services charge, and a few fee models to choose from.

If the data says shielding hurts turnout in some cases, that goes in the report too. Honestly, a mixed result is probably the most useful one for ENS.

You’ll see the plan, the first dataset and some early numbers here by 30 November, and the full report, ENS draft and memo by 21 December.

One change since I applied: SEEDGov asked @wyszomirski and me to coordinate, so he’s taking Arbitrum and I’ve dropped it from my scope. That brought my ask down from $2,000 to $1,700, and the dates now follow the new calendar.

Happy to answer anything here before the vote.

1 Like