RelicRouter: An SDK for Cozy Web Virality

RelicRouter: An SDK for Cozy Web Virality

by Matthew McDowell-Sweet (@msms) Miodrag Vujkovic (@miki)

A software development kit for apps that use rules to route blog posts, essays, music, podcasts, video, locations, products, resources etc. (aka “relics”) discovered online to relevant destinations.

Existing relic routing protocols lean on clunky toolstacks and high friction processes. Examples include:

  • Regular email digests of interesting artefacts
  • Targeted links sent to one person in the wake of a conversation
  • The depositing of a relic to a repo for later reference or use
  • Posting an on-brand thing to a group chat or private channel

Worse, current protocols exacerbate “route rot”—the phenomenon where the initial intent to share a discovered relic with someone, knowing it will resonate, decays over time due to the friction and unsuitable affordances involved in the act of sharing. As these intentions fade, relic circulation is artificially constrained and relationships that could have deepened do not.

Our SDK, in contrast, will enable apps that protocolise these relic routing patterns and nurture emerging/existing relationships. It will allow apps with end user workflows that:

  • Simplify relic capture, triage and annotation
  • Support routing at different tempos/volumes and for distinct segments
  • Output relics to any channel or medium
  • Permit both sender and recipient-side composability
  • Interop with and leverage ML and AI tech


What is the existing target protocol you are hoping to improve or enhance? Eg: hand-washing, traffic system, connector standards, carbon trading.

The sharing of something found online with someone/to somewhere that it is especially relevant.

What is the core idea or insight about potential improvement you want to pursue?

Routing something to a person or place whom you know will appreciate it should be as easy as realising its relevance in the first place. Right now, it is not.

People engage with blog posts, essays, music, podcasts, video, locations, products and resources they discover online. During engagement, the relevance of a thing to a particular individual, group or audience crosses an implicit threshold and is realised. It is is then annotated—either immediately or a short while later—with minimum viable context and distributed via whichever channel/medium best matches the communication context of the emerging/existing relationship.

People do this continuously, in spite of inherent friction in the processes and regardless of the lack of not-quite-fit-for-purpose tools.

What is your discovery methodology for investigating the current state of the target protocol? Eg: field observation, expert interviews, historical data analysis, failure event analysis

Product discovery methods (e.g. customer conversations, async continuous discovery) to understand the behaviours and systems people currently use to route relics to relevant destinations.

Cognitive task analysis to pursue a deeper comprehension of the cognition involved in discovering, engaging with, and routing relics to specific people and places.

In what form will you prototype your improvement idea? Eg: Code, reference design implementation, draft proposal shared with experts for feedback, A/B test of ideas with a test audience, prototype hardware, etc.

We will select key end user workflows that the apps based on our SDK would enable—e.g. send a weekly email digest, post things with tag X and Y to a specific Discord channel—and prototype them in collaboration with individuals engaging in that behaviour. For example, assembling a bespoke iOS app to generate and send an email digest, building a custom Discord bot to capture a relic and post it.

How will you field-test your improvement idea? Eg: run a restricted pilot at an event, simulation, workshop, etc.

We will involve app developers in the formulation of the SDK (e.g. the priority output channels, the most useful ML/AI interops, the mechanics for sender/receiver composability) and iterate upon prototype workflows that we surface to potential end users.

Who will be able to judge the quality of your output? Ideally name a few suitable judges.

Developers/engineers that want to build applications that use rules to route relics to relevant destinations.

How will you publish and evangelize your improvement idea? Eg: Submit proposal to a standards body, publish open-source code, produce and release a software development kit etc.

We will release a software development kit that enables the development of apps that protocolise relic routing patterns. This will be used to build novel products and tooling for sharing and circulating interesting things in a targeted way. We will also synthesise conducted interviews, analyses and prototyping efforts into longform textual output that references the SDK and contextualises its emergence.

What is the success vision for your idea?

We build the SDK that solves curation for the era of the Cozy Web, Dark Forest and Extended Internet Universe. Envisioned use cases and outcomes downstream of the RelicRouter SDK:

  • Consume and annotate a continuous queue of relics and distribute them in digest format at a maximum rate of two relics per day
  • Subscribe to a person’s output and elect to receive a monthly XMTP digest containing only relics with tags A, B, C and D
  • Post 5 queued items to a token-gated Warpcast group chat every three days and use any relic interactions to inform future 1-to-1 routing recommendations
  • Automatically send everything I annotate with tag X and categories Y and Z to a colleague via WhatsApp
  • Parse Zoom audio locally, surfacing real-time recommendations from my relic backlog, and share relics I approve with attendees the following day via email
  • Submit a podcast show notes URL and indicate the relics I would be most likely to explore and route to my top contacts based on my past behaviour
  • Construct a new standard for accurate, aggregated and anonymous relic analytics based on relic routing, zero knowledge and content addressing tech
  • Relics that achieve virality autonomously (think semantic samizdat) because one localised resonance pattern matches with other group’s/community’s relic-resonance profiles

This is great!

I’m wondering, do you think the rather efficient blockchain infrastructure can be used in some way?
Say someone connects their wallet and can now route incoming links to platforms of their liking irrespective of where the sender sent it?

Zapier did this and did it well, I’m wondering if web3 can be more efficient and democratized cause of the wallet connect.

Yes, blockchain infra of different flavours is definitely on the radar.

  1. We’d initially indexed on email as a proven, Lindy channel for distribution. Because the intent is to be channel agnostic (for outputs, especially), solving for wallet-based comms (likely via XMPT) seems like a sensible next step as its one of the newest kids on the block. Even better, the sender and receiver-side composability imagined for email (e.g. I send out an email with one thing from seven categories each week but a receiver is only interested in five categories and wants a monthly digest) is probably easier to iterate towards within wallets and token-based systems.

  2. Another possibility (admittedly more speculative right now) is leaning into the emergence of frames and actions on Farcaster. We haven’t done a deep dive, but a sufficiently capable RelicRouter could enable a lot of novel behaviours from the comfort of one’s own feed whilst still allowing for channel-specific conversations and value to accrue.

  3. Content addressing (e.g. Arweave, IPFS) could be really meaningful when combined with RelicRouter. Imagine that you and I, quite serendipitously, listen to the same podcast episode and route it to a different set of contacts or destinations. The podcast’s link/hash/address is unique and gets shared across two otherwise distinct networks in an identifiable way that is not beholden to centralisation under a typical platform. This matters and could lead to new ways of A) understanding how a single thing moves and flows amongst disparate networks and B) enabling relics to “island hop” based on relevance and in spite of balkanised, cozy web closed social dynamics.

1 Like

Open to suggestions for alternative terms to “relic”. “Artefact” has been used before by us but it is more obscure. “Thing” and “item” are vague and bland. “Link” is almost derogatory and strips the things people find of their context. The ideal term needs to communicate that it is a link that represents something which is loaded with context and meaning for the person finding it and for the people/place that person thinks it would resonate for.

1 Like


Clip (of CLIP fame)

1 Like

So the elephant question in the room is What’s the one thing the SDK do that Notion can’t?

And the elephant answer to that is Notion formats are already suprastructures or downstream of capture, maintenance and routing protocols. So an mvp on any one of those protocols sounds great.

I’d love to see the mvp really saving my thinking brain. If people need to actively think about capture or maintenance or routing, it’s not infra enough. It’ll be another Obsidian, Roam, Logseq … quoting xkcd:

Circumventing that is the challenge i think.


Check out the following sections from my local newspaper which I find rather enjoyable and which provide inspiration for this neat little protocol idea:

1 Like

I like “gem” and “extract” :smiley:

Notion ultimately wants you to use Notion more and more. But if I find something and think you’ll like it, that intent plus a little context should be enough to get it to you. Everything in between is irrelevant.

Kinda like how Stripe doesn’t really care about where a transaction happens, the finance/accounting package used by the vendor etc. They’re increasing the GDP of the internet by unblocking capital’s ability to flow and circulate. There’s currently no analogous toolset for unblocking the flow and circulation of “things” online. Trad. platforms worked for a while but we’re seeing their limits/impacts now.

That being said, what you prefer to use to interact and communicate—and what I use to do the same—matters. There’s an element of promise theory here, a need for composability for both the sending and the receiving party’s preferences.

Re: Obsidian, LogSeq etc: they assume the eventual intent is depth. And it may be for the individual. But for the P2P sharing and curation occurring between parties, depth is the enemy.

1 Like

Oh, cool! :pray: classic example use case. Things happen in a place; they get logged; the things that happened get sliced off the queue and shared with those who are interested.

1 Like

Absolutely! I actually think there’s a Steve Case style third wave of the web space for a nice blend between these types of super classic local community newsletter and web3 stuff. The CV Weekly has both a print edition that’s passed out at civic institutions and delivered the old school way. There’s also a website that sometimes has even more longer articles. They also literally do some Associated Press type stuff reprinting other things from other places. Anyway I have a loose ambition to sync up with Robin their publisher and talk about a pioneering spirit / future of the foothills style semi regular column. They have a long running “treasures of the valley” on history there. Anyway that might be a fun space to align with your energies here on experimenting with this type of cozy web sharing. Thanks for reading. Cheers, PA


connects to Warpcast