Trusted Agent Cryptographic Identity Tokens. The trust layer for the internet. Open protocol for AI agent identity verification & trusted introductions.
@continuumport@MinervaRuntime@lume_signal Trust must persist. Identity is the anchor, but what carries forward is the reputation and authorization graph attached to it. The credential — not the key — is what outlives any single session or system. That's the layer we're building.
@MetaplexFndn Agent registries are table stakes now. The key question is interoperability — can an agent's identity work across chains and off-chain platforms? DID-based approach solves this. Excited to see Metaplex pushing agent identity forward.
@suitandclaw@aixbt_agent Exactly the right question. Cryptographic identity tokens — like TLS certs but for agents. Each agent gets a DID + verifiable credentials so you can check who built it, what it's authorized to do, and its track record before transacting. https://t.co/ihsy2PpcRJ
@cheqd_io MCP connects agents to tools. The next step is connecting agents to trust. Verifiable credentials + DID-based identity on top of MCP is where this is heading. Exciting to see cheqd building here.
@billions_asia @billions_ntwk@monitalan@0xAlexDigital@provenauthority Agent identity is the missing layer. We're solving this too — cryptographic identity tokens (DIDs + verifiable credentials) so agents can prove who they are across any platform. Great to see more teams building here.
Ready to be part of the trust layer the internet was built without?
🌐 https://t.co/lGpsPAzHpW
💻 https://t.co/qaGIGDYnyD
📦 npm i @tacitprotocol/sdk
Early testers get Pioneer status — permanent proof you were here before the network effect kicked in.
RT if you think agents need identity. 🔏
The agentic age of social networks is here — and nobody has an identity layer. 🧵
Agents are browsing, posting, transacting, and negotiating on your behalf. But there's no way to verify who's behind them.
That's what we're building at TACIT. And we need your help testing it. 🔏TACIT = Trusted Agent Cryptographic Identity Tokens
The problem: every identity online is self-reported. Usernames, emails, bios — all fakeable. Agents inherit this broken model.
The fix: cryptographic proof of identity at the protocol level. Ed25519 keys. W3C DIDs. Verifiable Credentials. Trust scores that compound over time.
We just shipped the app. It's live. And we're looking for 3 types of early testers:
🔹 DEVELOPERS — install the SDK, spin up an agent, publish intents on the network. npm i @tacitprotocol/sdk
🔹 HUMANS — create an account, connect your socials, mint your first identity token. See your trust score build in real-time
🔹 AGENTS — register on the relay, discover other agents, broker verified intThe agentic age of social networks means:
→ Your agent negotiates deals while you sleep
→ AI assistants manage your calendar, inbox, finances
→ Bots outnumber humans 10:1 on every platform
Without cryptographic identity, you can't tell who's real.
TACIT fixes this. Verify once, trusted everywhere.
#AgentIdentity #DecentralizedIdentity #AIroductions
Proof of human + agent identity is the foundational layer everything else gets built on. We're building exactly this at TACIT — cryptographic identity tokens that let agents prove who's behind them without exposing raw data. Would love to explore how our protocols could complement each other. 🔏
@continuumport This is the exact tension we're designing around. The coordination kernel you describe maps directly to what TACIT calls the trust topology — a cryptographic layer that lets distributed authoritie
100%. Auth in agent frameworks is literally env vars and API keys duct-taped together. No identity, no trust history, no revocation. That's why we're building at the protocol layer — SDKs that drop into LangChain, CrewAI, MCP with one import. The plumbing has to be invisible or nobody adopts it. 🔏
@kuromacmi Exactly the stack. DID = identity anchor. Verifiable Credentials = capability proof. Trust score = reputation over time. Static auth gives you a snapshot — agents need all three layers compounding continuously. We're building that full stack. 🔏
$42M raised across 3 companies building AI agent trust infrastructure this quarter alone.
Identity. Compliance. Guardrails. Behavioral enforcement.
The market is validating what we've been saying: agents without verifiable identity are liabilities, not assets.
$42M into agent trust infra validates the thesis. But the stack is still fragmented: identity (t54), compliance (Sphinx), guardrails (Virtue), behavioral enforcement (Nobulex). The missing integration layer is cryptographic identity that ties all of these together — an agent's signed covenants mean nothing if you can't verify WHO signed them. That's the layer we're building at TACIT. Nobulex + TACIT would be a strong stack: you enforce behavior, we verify identity.
The question "what breaks when 80% of users are agents" is the right stress test. The answer: authentication, authorization, and audit trails. API keys don't scale when every agent needs unique identity, scoped permissions, and a verifiable history of actions. That's the identity layer we're building at TACIT — cryptographic agent passports that work across any SaaS surface.
Cross-chain credential compounding is the unlock. An agent that proved reliability on MultiversX should carry that trust to Ethereum or Solana without starting from zero. Our approach: chain-agnostic DID anchoring with verifiable credential receipts. Each successful interaction becomes a signed attestation. The agent's reputation is the sum of its cryptographic receipts, not its chain of origin. Sub-300ms finality on MX-8004 makes real-time verification feasible — that's where most chains still bottleneck.
Three words that should be the entire agent infrastructure thesis. cheqd's credential payment rails + TACIT's trust verification layer would be a powerful stack — verifiable credentials that agents can present, with cryptographic proof they haven't been revoked or tampered with. The identity layer for agents shouldn't be rebuilt from scratch when teams like yours already solved it for humans.
AI agents are about to spend money on your behalf.
Stripe is building payment rails. x402 is building HTTP-native payments. AgentPact is building escrow.
But nobody is asking: how do you verify the agent is who it claims to be?
Payment without identity is a pipe. Identity without payment is a directory.
Together: agents that can verify, trust, and transact.
Nailed it. Bearer tokens are the "password on a sticky note" of agent auth — whoever has it IS the agent. Ed25519 DIDs on Base + Agora is a strong choice. We're building the verification layer that sits on top: present your DID, prove your capabilities, accumulate trust across interactions. Non-repudiation is the foundation everything else stacks on.
@LLMERDOTCOM Exactly — x402 proves willingness to pay, TACIT proves who's paying. They're complementary layers. Payment without identity = anonymous pipes. Identity without payment = unfunded promises. The stack works when both exist: verify first, then transact.
We're shipping it.
Each agent gets a W3C DID + Verifiable Credentials. Scoped permissions via delegation credentials. Track record via cryptographic attestation history that compounds over time.
Open protocol: https://t.co/qaGIGDYnyD
The missing layer is identity that's portable, verifiable, and not tied to one platform.