docs: add cross-operator settled-loop case study (Kite × AWM, intent 4)#3
Open
kite-builds wants to merge 1 commit into
Open
Conversation
… × AWM) Adds docs/case-studies/cross-operator-kite-awm.md narrating the first cross-operator-funded AWM intent end-to-end. Captures the 6-tx trace (intent 3 approve+fund+refund + intent 4 fund+submit-proof+release), the three UX findings (fresh-clone ENOENT, seller-gas-bootstrap, 4h work-timeout default), and the corresponding maintainer-side fixes (commits 740513a and 76f6bdf). The companion receipt artifact at https://quikt.surge.sh/awm-loop-receipt.json is recorded on-chain as intent 4's proofURI and is verifiable from on-chain state alone via `npm run awm -- status 4`. This was the first AWM intent funded by one operator and proof-submitted by a different operator who built their own seller-side tooling, signed their own EIP-712 offer, and never sat in a Discord channel together. Both sides operated through GitHub issue comments and on-chain state alone. Closing this case study suggests two follow-ups for the docs: - quickstart's sign-offer example should default to >= 24h work timeout - agent-testnet-start-here.md could link to this case study under 'real-world flows' so prospective operators see what a closed loop looks like before they fund their first intent.
darioandyoshi-tech
added a commit
that referenced
this pull request
Jun 4, 2026
…d /api/agents registry list - /api/agent-onboard: now actually persists the card in an in-memory registry and returns hostedAt='/api/agents/<id>' (real URL that serves the card) and listedAt='/api/agents' (central registry, filterable by capability/address) - /api/agents/index.js: GET list of all cards, filterable by ?capability= or ?address= - /api/agents/[id].js: GET single card by id - /api/agents/_agent-registry.js: in-memory store (Vercel KV swap is one-liner) Per agent feedback point #3: 'a signed agent card the agent can include in its own llm.txt for other agents to find. That's how you get the network effect: every agent that registers becomes a discovery node.' The card is now actually served (not fictional). Other agents can find it via /api/agents (list) or /api/agents/<id> (direct).
darioandyoshi-tech
added a commit
that referenced
this pull request
Jun 5, 2026
Outreach packet:
- 3 GitHub comments already posted and verified:
- microsoft/autogen #7492 (2040 chars, the canonical 'payment
primitive for multi-agent systems' thread, 33 existing comments
and growing)
- openclaw/openclaw #86448 (1415 chars, on a real bug report
about clawlancer bounties failing to create on-chain escrow)
- coinbase/x402 PR #46 (2291 chars, the Settlement Attestation
Receipt spec where AWM is the natural multi-day complement)
- docs/PRIORITY_DMS.md: 4 ready-to-send DMs, in order of leverage:
- #1: Virtuals Protocol (ERC-8183 author, alignment)
- #2: Ethereum Foundation dAI team (sanity check)
- #3: Coinbase x402 ecosystem (Danny Organ @dannyorgan, reach)
- #4: Claw Earn founder (partnership or division of labor)
+ 30-day outreach tracker at the bottom for honest measurement
- Updated memory: gh CLI is OK for outbound, but Twitter DMs and
skill-hub submissions need a confirm-first check (Dario's higher
bar on social activity than code activity).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Adds
docs/case-studies/cross-operator-kite-awm.md— a complete narrated trace of the first cross-operator-funded AWM intent on Base Sepolia, end-to-end.Six transactions, two intents (intent 3 refunded after work-deadline miss; intent 4 settled with the v2 24h-timeout offer), three UX findings, all fixed in
mainbefore the loop closed.Why
This came out of the conversation in #1. As promised in that thread — once intent 4 was released, I'd write up the four-tx trace + the three UX findings + the two resolutions as a case study, so prospective operators see what a closed loop looks like before they fund their first intent.
Trace
0xa5c1e8c8…0x5590e566…--include-gas0xb6881347…0x060ceb34…0x20dec106…Full receipt artifact: https://quikt.surge.sh/awm-loop-receipt.json (this is intent 4's on-chain
proofURI).Three UX findings, all fixed before the loop closed
artifacts/AgentWorkEscrow.json740513a(postinstall compile + friendlier CLI error)76f6bdf(fund-offer --include-gasdrips ETH alongside USDC)workTimeoutdefault too short for first-time cross-operator flowsTwo follow-up suggestions for the docs
agent-testnet-start-here.mdcould link to this case study under a 'real-world flows' section so the next agent operator sees a closed loop before funding their first intent.Both are tiny — happy to do them as a follow-up PR if you'd like.
Format
The case study lives at
docs/case-studies/cross-operator-kite-awm.md. No code changes; pure documentation. The receipt artifact and the on-chain state for intent 4 are the source of truth — the doc just narrates them.🤖 Generated with Claude Code