KARTA — spend control for AI agents
Server Details
Task envelope, budget ledger, kill switch and signed receipts for agents that pay.
- Status
- Healthy
- Uptime
- 100.0% over the last 24 hours
- Last Tested
- Transport
- Streamable HTTP · MCP 2025-11-25
- URL
- Repository
- mrpkk/onchain-ai-agent
- GitHub Stars
- 0
- ChatGPT
- –
- Claude
- –
TDQS
Scored across 6 tools
Each tool targets a distinct action-resource pair: check_spend evaluates a proposal, get_envelope reads state, envelope_evidence produces a verifiable bundle, kill_envelope stops, open_envelope creates, and policy_rules explains the rule set. The descriptions make the boundaries between check_spend, get_envelope, and envelope_evidence explicit (verdict vs. state vs. evidence pack).
Four tools follow a clear verb_noun pattern (check_spend, get_envelope, kill_envelope, open_envelope), but envelope_evidence and policy_rules use noun-first naming. The deviation is minor and still readable, so it does not seriously hinder selection.
Six tools is well-scoped for a spend-control lifecycle: open, check, inspect, evidence, kill, and policy explanation. Every tool earns its place without bloat or thinness.
The core lifecycle is covered, but there is no way to list or enumerate envelopes, no explicit record_spend tool despite evidence returning recorded spends, and no update/adjust operation for an envelope. These are notable gaps that could force agents into dead ends when managing multiple envelopes.
Available Tools
6 toolscheck_spendCheck a spend against a task envelopeBRead-onlyIdempotentInspect
Check whether a proposed agent spend is inside a task envelope. Returns verdict, rule, reason, counterfactual and next. No state is stored.
Inputs: payee (required), amount (required), envelope (required), calls_so_far, +2 more. Returns: result
envelope_evidenceEvidence pack for an envelopeBRead-onlyIdempotentInspect
Return an evidence bundle for one envelope: state, every recorded spend, the stop record and a sha256 of the whole pack that anyone can recompute. Read-only.
Inputs: envelope_id (required). Returns: result
get_envelopeRead an envelopeARead-onlyIdempotentInspect
Return an envelope's state: budget, per-call cap, calls used, spent, remaining, allowlist and expiry. Read-only.
Inputs: envelope_id (required). Returns: result
kill_envelopeStop an envelopeADestructiveIdempotentInspect
Stop one task envelope immediately: no further spend is approved for it, the reason is recorded with a timestamp. Idempotent — the first reason wins.
Inputs: envelope_id (required), reason. Returns: result
open_envelopeOpen a task envelopeAInspect
Create a signed budget envelope for one task: total budget, per-call cap, call count, allowlisted payees, subscription rule, expiry. Stores it in this server's local ledger (no funds move, no keys). Returns the envelope id. Capped at 30 envelopes/hour per server.
Inputs: calls_max (required), budget_total (required), per_call_max (required), allowlist, +4 more. Returns: result
policy_rulesList policy rulesARead-onlyIdempotentInspect
List the deterministic rules a spend is checked against (E01-E08) with the order they are applied and the meaning of each. Read-only, no state.
Inputs: no named parameters. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Related MCP Connectors
Signed agent discovery, verified escrow funding, paid work, settlement and usefulness receipts.
Agent memory and a signed spending ledger: pay any x402 API, keep a provable receipt.
Bounded digital tasks, simple checkout, private result recovery and signed receipts.
Cryptographically anchored evidence for agents: verified run receipts, proof-gated settlement.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceGoverned robot task authority — signed mandates, bounded envelopes, evidence for every task.MIT
- AlicenseAqualityCmaintenanceEnables AI agents to create cryptographically verifiable receipts of their delegated work, with capabilities for multi-party approval and offline verification.1184 npmApache 2.0

Mandareofficial
AlicenseAqualityBmaintenanceAccountability for AI agent fleets: verify a signed, hash-chained Mandare ledger, check budgets, kill an agent, and issue passports and mandates.82AGPL 3.0
evermint-mcpofficial
AlicenseNot gradedqualityDmaintenanceTamper-evident receipts for AI agent actions. The notary layer for agent-to-agent transactions.40 npm1MIT
Glama MCP Gateway
Add one secure layer between your agents and this server.