Vendor: Cyan Eclipse Technologies · Platforms: MCP over stdio
Certification scope
Test Bench v0.1.0, 2026-10-24, sandbox environment: policy enforcement (no-policy/over-cap/off-allowlist denied), approval queue, kill switch, append-only audit, schema and error handling — 12 pass, 1 na of 13 checks
What it does
- Create agent wallets owned by a named human
- Grant written spending policy: per-transaction cap, daily cap, recipient allowlist, approval threshold, expiry
- Dry-run payment check against policy (allow / approve / deny with reason)
- Execute payments inside the granted scope (sandbox simulated USDC)
- Human approval queue for over-threshold payments
- Instant kill switch (revoke) freezing all spending
- Append-only audit log of grants, payments, approvals, revocations
Tools
create_walletget_walletset_policyfund_walletcheck_paymentpayapprove_paymentdeny_paymentaudit_logrevoke
Problems it addresses
- I need to give an SI agent spending limits so it can pay for things without spending freely
- My agent needs to make payments but only to approved recipients and only up to a cap
- I need a kill switch that stops my agent's spending immediately
- I need proof of exactly what my agent was authorized to spend, for liability
- Payments above a threshold should wait for a human approval before executing
Not tested
- Auth enforcement (bench U1 = na: local stdio transport has no credential layer to test)
- Production payment rails — bench ran against simulated test USDC only; x402/production rails not yet wired (x402_ready: false)
Bench scorecard: test-bench/results/eclipse-agent-wallet-mcp-2026-10-24-2.json
Evaluation-intake listing
This listing records the tool's place in the Bazaar's evaluation
and monitoring pipeline, with the evidence gathered so far printed above — verdict, scope,
and not-tested items included. It is not sold on this site, and its presence in the catalog
is not a claim beyond that evidence. The service behind every verdict is the bench:
what a bench run buys →