Media, Docs & Compass: Project Proposal [Vote Open]
_Final version of the proposal discussed here, now open for a Snapshot vote.
Status: FOR APPROVAL — Snapshot vote · framework: AIP-006 · a pass is recorded downstream as an amendment to AIP-007 (arrowair.eth)
Snapshot vote: opens Saturday 4 July 2026, runs one week, closes Saturday 11 July 2026. I am starting all three projects at my own risk on Monday 6 July 2026, ahead of the close (see Early start below). The term runs 28 weeks (7 four-week pay periods) from that date, with funding expiring ~18 January 2027 so the DAO actively reviews rather than auto-renews. If the vote does not pass, I receive nothing for the work done in the interim.
This document is context, not the contract. It sets out the reasoning and full detail behind the three projects to inform discussion and the vote. The binding terms (gates, caps, pro-ration) are frozen in each project’s repo at the
v1-proposaltag (social-comms · docs-improvements · arrow-compass) and recorded via the AIP-007 amendment. The vote itself is on Snapshot. This post can be edited; those frozen specs cannot.
Summary
This proposal puts three DAO-level projects to a Snapshot vote, all led by me (Sleety), together totalling ~32.5 hrs/week and a combined $8,760 per 4-week pay period across the three caps ($61,320 maximum across the full 7-period / 28-week term, released period-by-period only against delivery):
| Project | hrs/wk | Level (AIP-004) | Monthly cap |
|---|---|---|---|
| Media Updates & Comms | ~17.5 | L4 ($64) | $4,480 |
| Docs Maintainer & Improver | ~10 | L5 ($75) | $3,000 |
| Arrow Compass | ~5 | L4 ($64) | $1,280 |
| Total | ~32.5 | $8,760 |
They are three separate AIP-007 projects, each with its own cap, gate, repo, and accountability table, so the DAO can renew or wind down any one independently. I present them together because they reinforce each other: Compass sets the priorities, Media broadcasts progress against them, and Docs makes sure anyone the media pulls in can actually get started. This is the vote-ready version of the forum discussion, with the open parameters now locked.
All three run on the same principle: gated, paid on delivery. Each defines near-binary payment criteria up front, clear enough that a person or an agent can check whether they were met, and every period is logged in a public accountability table. Compensation only releases against work that shipped, so the DAO is never paying for intention instead of output. I volunteer these as the first pilot of this model, which means they come in on or below budget by construction.
The pay-on-delivery model
We took a bigger swing at comms last year with Project Visibility, and it never reached execution, partly because nothing tied payment to delivery. These three are built the opposite way.
A note on periods: throughout this proposal a “month” means a 4-week pay period, following AIP-004’s 48-week year of twelve 4-week periods. Every cap below is per 4-week period, and compensation disburses at the end of each one. Each project keeps a single, public accountability table in its repo, updated every pay period: each gated deliverable, what was committed, what was delivered, whether the floor was met, and links to the actual work. Before any payment releases, the table is the thing checked. If there is no table, funds are not released. Where a target is missed, only that line pro-rates, on the fixed weightings locked below, so a strong week on the easy metric can never rescue a weak one on the hard metric.
This keeps the lead honest, gives the DAO a clean audit trail, and means nobody has to take anyone’s word for anything. I would like this to become the default for how we run projects and bounties going forward.
Early start, at the leader’s risk
I am not waiting for the vote to close before beginning. Since there was no objections raised in the DAO forum project proposal draft, I am starting all three on Monday 6 July 2026 while the vote runs. This is deliberately at my own risk: the first accountability period runs from 6 July, and if the vote does not pass I accept that I would receive no compensation for the work done in that window.
If it passes, the first month’s cap covers work from 6 July, so ratification confirms funding for output already visible in the accountability tables. Practically, this asks the GBC to treat 6 July as the funding-eligibility start once the amendment ratifies. It fits the spirit of the whole proposal: the DAO only ever pays for output it can already see.
1. Media Updates & Comms
Purpose
To consistently document and broadcast Arrow’s work in public: recruiting contributors, spreading our bounties, and building credibility by turning the everyday progress of an open aircraft hardware ecosystem into a steady, interesting stream of media.
Scope
A reliable content stream from the official accounts: build footage, CAD and prototype shots, screencasts, memes and GIFs, test clips, audio, plus being genuinely present in the replies instead of posting into the void. We have multiple aircraft moving through design, prototyping, and testing, and almost none of that work gets published today. As a build-in-public org we owe it to the contributors who have not found us yet to show that work louder than we would if we were only building for ourselves. It is also our route to investors, collaborators, suppliers, workshops, media outlets, and co-marketing partners.
The formula is simple: interesting media, then a link to get involved, then a call to action, repeated daily. That ties straight to AIP-008: an open, internet-native org only grows if people can see it happening.
The spirit is regular-cadence, short-form content, the steady drip that keeps Arrow visible, not big documentary productions. Channels are evolving, but right now it is mostly X/Twitter, where a couple of weeks of testing has worked well, with YouTube, LinkedIn, Instagram, and TikTok coming in over time as downstream channels. The proposal deliberately does not lock the platform mix; I will monitor analytics and follow what works, favouring consistency and impact on one platform first, then piping the same content to others downstream.
Not in scope: long-form, heavily-edited video. That is high-effort work, and meeting a publication cadence would force me to rush or compromise it. When a big edit is worth doing it is better rewarded on its own merits through a retroactive grant or a standalone bounty, not folded into this project where it would quietly reduce output to a couple of posts a month.
Project lead
Me (Sleety). Media and content is my main area of expertise, and I understand Arrow’s culture, brand, and voice from the inside. In anticipation of this project I ran an experiment between May and June 2026: 10 posts and 36 replies, which brought 75 new followers on X/Twitter, 341 likes, 326 profile visits, 822 engagements, and over 24,000 impressions. I enjoyed it, and it is a lot of work that needs daily creativity and commitment to do well.
Payment gate
A weekly threshold of short-form content, set against the comp being asked for:
- ≥ 7 short-form posts/clips per week across our active channels (primarily X today), and
- ≥ 3 substantive replies/comments per day, or ≥ 15 per week, for genuine engagement presence rather than pure broadcasting.
Posts and replies are scored separately, each against its own share of the cap. The cap is split 70% posts / 30% replies = $3,136 / $1,344.
Locked pro-ration:
- Posts ($3,136): the posts share is $784 per week. Meeting the weekly floor of 7 releases that week’s posts share in full. Falling short docks 1/7 of that week’s share per missing post, so a week that hits 6 of 7 releases ~$672 and docks ~$112.
- Replies ($1,344): the replies share is $336 per week, released on meeting the weekly floor, and pro-rated proportionally to any shortfall against it.
These are floors, not ceilings. The cap pays for the role and the booked hours, not per post, so 7 is the minimum that releases the month, not a number to stop at.
Worked example (accountability table, updated weekly in the repo):
| Week | Posts/clips (≥ 7) | Posts met? | Replies/day (≥ 3) | Replies met? | Evidence |
|---|---|---|---|---|---|
| Wk 1 | 9 |  |
3.4 avg | |
links |
| Wk 2 | 8 | |
3.1 avg | |
links |
| Wk 3 | 6 |  86% |
3.0 avg | |
links |
| Wk 4 | 10 | |
3.5 avg | |
links |
Replies cleared every week, so the full $1,344 releases. Posts cleared three weeks; week 3 hit 6 of 7, docking ~$112, so the posts share releases $3,024 of $3,136. Month total: $4,368 of $4,480, and the strong reply weeks did nothing to rescue the posts miss.
Cost
~17.5 hrs/wk = 70 hrs/mo at Level 4 ($64), so the monthly cap is $4,480, gated monthly, no rollover. That is pure labour (research, composing, editing, scheduling); I already hold several thousand dollars of tooling (Adobe Suite, Cinema 4D, Claude, ChatGPT, Midjourney, and more) and absorb any hours beyond the cap. Level 4 reflects a mix of highly skilled media work and moderately skilled marketing and comms.
2. Docs Maintainer & Improver
Purpose
To treat Arrow’s documentation as a living product: keeping the global docs experience clear, beautiful, and low-friction as the DAO grows, so any contributor anywhere can find what they need and get started.
Scope
This is the ongoing stewardship layer for the docs platform that Project Onboarding built, not a rebuild. Onboarding stood up the website, the Docusaurus docs site, and the first guides as a one-time effort, and it has since expired. But docs are an infinite game, and right now keeping them current is nobody’s job, so they drift.
This role owns global docs infrastructure: structure, styling, UX, new features, and the guides that should appear whenever something ships or changes. Individual project docs stay with their teams; this is the layer underneath them. It is the most direct expression of AIP-008’s minimal-friction goal for global contributors, and it deliberately does not overlap the completed Onboarding build.
Project lead
Me (Sleety). I designed and built a significant portion of the current docs platform, so the context is already there. I have ambitions for what our docs could be, and I will stay pragmatic about shipping incremental improvements alongside the bigger experiments.
Payment gate
A monthly floor of shipped, live-and-merged work, split across the role’s two jobs, improve and maintain. The cap is split evenly four ways at $750 per line:
- Copy improvements ($750): ≥ 8 per month (rewrites, new guides, clarity passes, restructures). Pro-rates at ~$93.75 per item short.
- Functionality improvements ($750): ≥ 4 per month (an integrated
.mdxdoc-app, an embed, an automation, a new interactive feature). Pro-rates at ~$187.50 per item short. - Maintenance kept current ($750): open docs issues (broken links, outdated pages, doc bug reports) triaged and resolved within the month. Pro-rates by issues resolved against issues open.
- New-feature coverage ($750): whenever something ships in the DAO (a feature, a passed AIP, a process change), a corresponding doc or guide exists within ~7 days. Pro-rates by features covered against features shipped.
Every item is a real artifact an agent can verify is live in the repo or site.
Worked example (accountability table, updated monthly in the repo):
| Criterion | Committed | Delivered | Met? | Line value | Released | Evidence |
|---|---|---|---|---|---|---|
| Copy improvements | ≥ 8 | 9 | |
$750 | $750 | links |
| Functionality improvements | ≥ 4 | 3 | 75% |
$750 | $562.50 | links |
| Maintenance (issues current) | backlog cleared | 7 of 7 closed | |
$750 | $750 | issue links |
| New-feature coverage | doc within ~7 days | 2 of 2 features | |
$750 | $750 | doc links |
Three lines clear the threshold; functionality lands at 3 of 4, docking ~$187.50, so the month releases $2,812.50 of $3,000. Paid for what shipped, every line checkable in the website repo.
Cost
~10 hrs/wk = 40 hrs/mo at Level 5 ($75), so the monthly cap is $3,000. Level 5 reflects the highest-quality deliverables and specialist knowledge of our docs, code, design, and infrastructure.
3. Arrow Compass
Purpose
To give the whole DAO a single, visible set of community-voted priorities each quarter, plus the orchestration to turn them into real bounties and progress, so anyone who wants to move Arrow forward knows exactly where to push.
Scope
A community-voted set of quarterly priorities, locked in via Snapshot, that act as a totem anyone can check. Two weeks before each new quarter we hold an open working call to debate and pick the goals that matter most. Then someone (me) orchestrates it: spinning those goals into bounties, running the cadence, and reporting honestly on whether we are getting closer. The priorities are selected to make the AIP-008 roadmap concrete one quarter at a time, and are expected to sit mostly at the DAO level.
Each quarter produces two tiers of priority, with me as Keeper of the Compass:
- 3 push goals: the quarter’s real focus, where we want measurable forward progress. Occasionally 2 or 4 if a quarter genuinely calls for it, but 3 is the target.
- 2–3 “don’t-drop-the-ball” priorities: the steady-state commitments that must keep ticking and must not regress while everyone chases the push goals.
Plus the orchestration around them: a meeting and brainstorm every two weeks dedicated to creating bounties that move the push goals, and an honest progress update published every two weeks (are we closer on the three, and are the steady commitments still healthy?), with wins and losses shared openly so others can step up.
Project lead
Me (Sleety). The Keeper is a facilitator, not an executor. The Keeper is not on the hook for single-handedly delivering the priorities; that is unrealistic and not the point. The job is the things a facilitator can control: running the cadence, creating and assigning bounties, encouraging participation, and documenting movement. The community does the building; the Keeper keeps it pointed and visible.
Payment gate
Gated on facilitation, not outcomes, all verifiable artifacts. The cap adapts to the month’s structure:
- Standard month (cap split 3 ways, ~$426.67 per line): both meetings held (one every two weeks), both progress updates posted, and ≥ 4 bounties created or assigned against the live priorities. Meetings pro-rate at $213.33 per meeting, updates at $213.33 per update, bounties at ~$106.67 per bounty short of 4.
- Quarter-start month (cap split 4 ways, $320 per line): the three lines above plus a fourth, the Snapshot vote held and priorities published (3 push goals + 2–3 don’t-drop-the-ball items). Same $1,280 cap, just weighted across the extra quarterly deliverable.
The Keeper is measured on initiating and encouraging the work, not on whether the DAO finishes it.
Worked example (standard mid-quarter month):
| Criterion | Committed | Delivered | Met? | Line value | Released | Evidence |
|---|---|---|---|---|---|---|
| Meetings held (every 2 weeks) | 2 | 2 | |
$426.67 | $426.67 | links |
| Progress update posted | 2 | 2 | |
$426.67 | $426.67 | links |
| Bounties created / assigned | ≥ 4 | 3 | 75% |
$426.66 | $320.00 | bounty links |
Everything clears except bounties at 3 of 4, docking ~$106.67, so the month releases ~$1,173 of $1,280. Because the gate is facilitation rather than outcomes, none of this depends on how far the DAO actually moved the priorities.
Cost
~5 hrs/wk = 20 hrs/mo at Level 4 ($64), so the monthly cap is $1,280, gated, no rollover. The Keeper’s responsibility is to identify ways to improve the DAO, coordinate discussion, create solutions through bounties, report honestly on progress, and follow up with contributors. Contribution is measured through the accountability table, not through outcomes the Keeper cannot control.
Alignment with AIP-008
AIP-008 sets Arrow’s mission as an open, internet-native hardware org that grows through global, low-friction contribution. These three are the reinforcing arms of that mission:
- Media is the visibility engine: an open org only grows if people can see it happening, so it broadcasts real progress and pulls new contributors in.
- Docs keeps the friction minimal for everyone the media pulls in, so a newcomer anywhere can find what they need and start.
- Compass turns the multi-year roadmap into a concrete, voted focus each quarter, so contribution points the same direction instead of scattering.
They also share a cadence: Compass’s fortnightly meeting is the natural home for Media (and any future Voice or Changelog work) to plug into, so we are not paying three times over for three separate meetings.
Uniqueness (AIP-006 gate)
Checked against the live AIP-007 projects list. The four active projects (Quiver, Spearhead, Caribou, Longshot) are all aircraft and hardware builds; none overlap these three. Two expired projects give the direct lineage: Docs is the pay-on-delivery continuation of the completed Project Onboarding, and Media is the pay-on-delivery successor to Project Visibility. Compass is a new coordination function with no existing equivalent. Each is a distinct, non-overlapping area of focus with its own leader, cap, expiration, and repo, which is exactly why they are presented as three AIP-007 rows rather than one.
How work gets done
I lead the three projects and report deliverables in each project’s accountability table. Broader execution is issued as grants and bounties, coordinated with the Grants and Bounties Committee (AIP-002), which holds a seat alongside the leader on one shared project multisig (per-project spend tracked in the accountability tables, rather than three separate wallets for a single lead). Per-period caps reset with no rollover. I will post narrative updates and sentiment polls to the forum per AIP-006, and negative sentiment is grounds to reconsider or wind a project down.
On what is being voted: the binding gates and pro-ration rules are frozen in the AIP-007 amendment PR, not in this post. This forum write-up is context and can be edited; the amendment PR is the fixed artifact the Snapshot ratifies, and any change to the gates after ratification needs a fresh Snapshot.
Midpoint review (week 14)
Halfway through the term, in the week of 12 October 2026 (week 14 of 28), each project gets a formal midpoint checkpoint so the community can reflect on whether it is worth continuing to fund. For each project I publish a midpoint report (the cumulative accountability tables, what shipped, spend to date, and an honest self-assessment against the gates), and each report runs its own community sentiment poll on the forum.
The trigger is mechanical, so nobody argues about what “negative” means on the day:
- I publish a per-project midpoint report and a 7-day sentiment poll opens on the forum for each project (continue / wind down).
- If a project’s poll closes net-negative (more “wind down” than “continue”, meeting a fixed 2M $ARROW quorum), a binding Snapshot halt-vote for that project opens within 7 days of the poll closing, opened by any GBC member or by me. The project halts on a simple majority in favour of halting.
- Projects whose polls do not close net-negative continue to the end of the term.
Because the three are independent AIP-007 entries, any one can be stopped this way while the others carry on. If a project is halted, its funding stops at the week-14 mark: the accountability table pro-rates the in-progress period’s cap to the work delivered up to that point, and no further periods are funded. Everything already shipped and logged is still paid; nothing beyond week 14 is.
Timeline
- Snapshot opens: Saturday 4 July 2026
- Snapshot closes: Saturday 11 July 2026 (one week)
- Work begins (at the leader’s risk): Monday 6 July 2026, ahead of the close
- Midpoint review (per project): week of 12 October 2026 (week 14 of 28), continue-by-default with a halt-vote off-ramp
- Term: 28 weeks (7 four-week pay periods) from 6 July, funding expiring ~18 January 2027, with an end-of-term review at the 28-week mark to renew or wind down.
Compensation is disbursed at the end of each 4-week pay period, once the accountability tables show that period’s criteria were met. The first period is contingent on the vote passing; if it does not, no compensation is owed for the at-risk work.
Snapshot vote
Question: Shall the DAO approve the three projects below (Media Updates & Comms, Docs Maintainer & Improver, Arrow Compass), all led by Sleety (@sl33ty), with a funding-eligibility start of 6 July 2026 (work begun at the leader’s risk ahead of this vote) for a 28-week term of 7 four-week pay periods (funding expiring ~18 January 2027), with a per-project midpoint continuation checkpoint at week 14 (~12 October 2026, continue-by-default with a halt-vote off-ramp), under the gated pay-on-delivery terms and per-period caps specified above (Media $4,480, Docs $3,000, Compass $1,280; combined $8,760 per 4-week period, each capped and gated independently)? A pass approves the projects and their funding on these terms; they are then officially recorded via an amendment to the AIP-007 projects list.
Why a single vote for three projects: they share one lead, one cadence (Compass’s fortnightly meeting is where Media and Docs plug in), and one pay-on-delivery model, so they are proposed and governed as a set rather than three disconnected asks. They stay independent AIP-007 entries, and the week-14 per-project midpoint checkpoint means a single For does not lock any one of them in for the full term. If you would rather these were voted separately, say so in the thread before the vote opens and I will split them.
Options:
- For — approve all three projects on the terms above.
- Against — do not approve.
- Abstain
A “For” result is the DAO’s approval of the projects. Recording them is the downstream step: a single amendment adds three rows (Media, Docs, Compass) to the AIP-007 projects list. See the companion aip-007-entry.md for the exact rows and the frozen binding gates.