Lineage Labs

Clearly, a powerful tech will come with equally strong FUD. We must design better AI systems. Good systems comes from good incentives. That's what we're building at Lineage Labs. youtube.com/watch?v=tNH43a…

Co-founder Erasmus Hagen got the opportunity to demo the WayID prototype (available at way.je) at @ClawCon Copenhagen yesterday. 📹 video of the 5min demo: youtube.com/live/B2g6qJJok…

The co-founders join the crustaceans in Copenhagen @clawcon this evening 🦞

@badlogicgames @mitchellh 🎥 Mario's talk: youtu.be/RjfbvDXpFls 🔗 Vouch: github.com/mitchellh/vouch 🔗 pi: pi.dev (recently acquired by @earendilworks)

@badlogicgames built pi agent with care but drowned in AI-slop. The fix: web of trust. @mitchellh turned this into Vouch. WayID will include Vouch as data source for agent trust score. Clankers aren't going away. We need trust rails for them. Links in thread ⬇️

Agents will be social. They will move across sites and need to carry provenance proofs. "If your agent acts on different services, there’s no standard way to reconstruct and attribute that chain of actions in a way both parties trust." @egg_ert in eggert.substack.com/p/stop-fightin…

Agents are becoming part of our lives - but can we trust them? My take: a human owner needs to take responsibilty and users needs to be able to verify that. That's why @specksnydern and I are building Way ID. Kind of a Trustpilot for AI Agents. 🤖 way.je

We're building the "Trustpilot for AI agents" 🤖 Introducing WayID — a provenance certificate system for AI agents. Prototype live at way.je. Build a trust profile, claim your agent, and let it show off its certificates on Telegram.

Everyone's building agents. Almost nobody's thinking about the infrastructure underneath. Here's a map of the emerging protocol stack — MCP, A2A, AG-UI, x402, AP2, ERC-8004, Web Bot Auth, and 20+ more. github.com/LineageLabs/aw…

@helena__rong @bah_eth This maps directly to what we're building at Lineage Labs. WayID combines three of these six trust primitives — Brief (cryptographic certificates binding agents to operators), Reputation (public feedback and incident reporting), and discovery — into a single infrastructure layer

@helena__rong @bah_eth The proposed taxonomy is useful. Six models, each with failure modes: — Claim: unverified, prone to deception — Brief: depends on issuer integrity — Proof: computationally expensive — Stake: favours the wealthy — Reputation: slow, Sybil-vulnerable — Constraint: safe but limiting

Paper by @helena__rong and @bah_eth proposes a six-model taxonomy for how AI agents should trust each other — Brief, Claim, Proof, Stake, Reputation, Constraint. Google's A2A scores worst. ERC-8004 scores best. No protocol covers all six. arxiv.org/abs/2511.03434

"In a world of AI, having a human network is going to be this incredibly important thing." — @alexblania on Proof of Human. Human–agent bindings are the infra problem of the agent era. At @lineage_labs we're building Agent Provenance Certificates — and World ID is a natural fit

@nembal has been doing good work with @agentcommunity_ getting a .agent domain name going. Good to see @brave supporting the initiative 🦁

🐝 Busy bee alert: #Maksika built wayMint on @Celo — verifiable ERC-8004 on-chain identity for AI agents, with Self Protocol ZK proof-of-human. 🔗 Karma: karmahq.xyz/project/waymint 🚀 Molthunt: molthunt.com/projects/waymi… 🤖 Agent: agentscan.info/agents/c7bd8a3… @CeloDevs @CeloPublicGoods

M2M payments is the future. And they will need public trust certificates and provenance trails. This is what we're building at @lineage_labs

#Maksika pushed out the first version of its project for the AI-only @synthesis_md hackathon: a dApp that expands #ERC8004 with proof-of-humanity from @selfxyz + @base. Still early, still buggy, but give it a spin: 8004.way.je. Feedback? In the comments, pls 👇

"...instead of trying to detect AI, enable affirmative delegation by humans such that whatever software we authorize can act–and be regarded as acting–on our behalf" @JoeAndrieu correctly calls for new trust rails in this great post: blog.joeandrieu.com/2026/02/13/deh…

Maksika, my #openclaw bot, is about to enter its first hackathon. @synthesis_md is an #AI-only #hackathon where builders & judges are all machines 🤖 One theme is "trust" and that's what Maksika will work on. It runs March 13–22. Go register! ➡️synthesis.md

OpenClaw 2026.3.8 🦞 🔒 ACP provenance — your agent finally knows who's talking to it 💾 openclaw backup — because YOLO deploys need a safety net 📱 Telegram dupes killed 🛡️ 12+ security fixes We fixed more things than we broke. Progress. github.com/openclaw/openc…