RAVEN. / for agentsRAVEN. / for agentsRaven is verification infrastructure built to be consumed by other agents. Machine-readable: /agents.json (capability manifest) and /openapi.json (API contract). Everything below is what your agent can do with it.
Give Raven a mint + token program; get a deterministic verdict (pass / pass_with_info_finding / warning / risk / unknowable) derived from decoded on-chain bytes: issuer authorities, full Token-2022 extension tail, metadata, and optional liquidity/venue evidence.
Every verdict ships with a domain-separated ed25519 signature over a canonical replay hash. Your agent can verify offline against the published key rvk_c2997e90215279c2 (GET /pubkey) and prove later exactly what Raven said, about which mint, at which slot. Verification recipe →
Responses include coverageGaps: the surfaces NOT evaluated (e.g. liquidity, top_holders, deployer_outcomes). Raven fails closed: missing evidence becomes explicit findings or unknowable — never a silent pass. Your agent can make its own risk policy from honest inputs.
npx -y raven-verify-mcp
One tool: raven_verify_token. Quickstart →
curl -X POST https://raven-hosted-verifier.onrender.com/verify \
-H "x-api-key: $KEY" -H "content-type: application/json" \
-d '{"mintAddress":"...","tokenProgramAddress":"..."}'
Invite-gated alpha; 10 req/min per key. Contract: /openapi.json. Request a key →
RAVEN is a registered ACP provider. Offering solana_verify_token at 0.1 USDC per job: escrowed payment, signed deliverable, on-chain settlement. No API key needed — your agent pays per verification.
Raven does not replace your agent, wallet, launchpad, or compliance process. It gives them a signed, deterministic evidence receipt before they touch a Solana token — your system keeps control of the final decision. Harness, not hype: the value is the deterministic verdict path, receipt verification, coverage gaps, evals, and policy handoff. Cheap deterministic preflight: repeated evidence checks should not burn expensive reasoning tokens — Raven answers in milliseconds for a fraction of a cent, no LLM in the verdict path. Open posture: MCP, ACP, OpenAPI, and AgentKit-style plugins — with no wallet signer access required, ever. Pilot in a week: readiness checklist.
Whether you call it an agent, bot, workflow, script, launchpad, backend job, or human review console — the contract is identical: call Raven, verify the signed receipt, inspect verdict/findings/coverage gaps, apply your policy, store the exact receipt, re-verify before material delayed action. The receipt is the portable artifact; context is not evidence; your policy owns the final decision — Raven never makes the business call.
token candidate -> collect mint/program (+metadata/pool if known) -> call Raven -> deterministic engine checks evidence -> signed receipt -> verify ed25519 vs /pubkey -> check verdict/findings/gaps/staleness -> apply YOUR policy -> store exact receipt -> proceed / block / escalate / reverify
Prohibited sequence: agent sees token -> asks an LLM "is this safe?" -> LLM reads social links -> LLM says safe -> agent acts. No receipt, no signature, no gaps — that is slop, not verification (anti-slop policy).
Raven is not a model wrapper. It is a verification harness: deterministic checks, signed receipts, coverage gaps, evals, public-key verification, fail-closed policy. Models and agent UIs change; the signed receipt contract is the stable artifact. Drill it before you trust it: failure drills.
Raven is a signed evidence-and-receipt layer for AI agents before they touch a Solana token. It verifies token-launch state from decoded on-chain bytes and returns a deterministic verdict with finding codes, explicit coverage gaps, and an ed25519 signature (key id rvk_c2997e90215279c2).
Before trading, listing, recommending, or settling around a Solana token — any moment where the agent needs verifiable launch-state evidence rather than a reputation score. Typical pattern: verify, store the receipt, proceed only on pass-grade verdicts under the agent's own gap policy.
Required: mintAddress and tokenProgramAddress. Optional: metadataAddress, poolAddress (Raydium CPMM, adds liquidity/vault evidence), commitment. Raven never accepts caller infrastructure: rpcUrl and issuerIdentity are rejected.
One verdict — pass | pass_with_info_finding | warning | risk | unknowable — plus findingCodes, findings with evidence, coverageGaps (surfaces NOT evaluated), engineVersion, observed slot, replayHash, officialAttestationHash, keyId, and an ed25519 signature. The whole response is a machine-verifiable receipt.
Raven signs every verdict, lists what it did NOT check, fails closed instead of guessing, tracks engine versions so history is never silently rewritten, and produces replayable receipts anyone can verify offline. It publishes findings, not a score.
Raven never claims a token is "safe", never predicts price, never gives financial advice, and never presents unverified surfaces as fine. With coverage gaps remaining, the strongest honest claim is: not enough evidence for a full pass.
Fetch GET /pubkey (key rvk_c2997e90215279c2, domain raven-official-attestation), recompute replayHash from the response fields via canonical JSON, then verify the ed25519 signature over domain || officialAttestationHash. Recipe: /security.html.
Three ways: free local MCP (npx -y raven-verify-mcp); hosted API key by request (free during invite-gated alpha) at /request-access.html; or agent-to-agent via Virtuals ACP, offering solana_verify_token at 0.1 USDC per job with escrow and on-chain settlement — no key needed.