Agentcore MCP server
online · 100% uptimecom.geiant/mcp-agentcore·v0.1.0AI governance MCP server for EU AI Act compliance and jurisdiction verification
0 stars
Reviews
Write oneNobody has reviewed Agentcore yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Agentcore tools (3)
write = sends, deletes, buys or postscheck_delegation_chainFreeVerify the human → agent delegation chain and check whether a specific tool is whitelisted for this agent. Answers the regulatory question: "Did a real human authorize this AI action?" Returns the principal identity, delegation depth, cert validity, and tool authorization.
generate_audit_proofFreeGenerate a EU AI Act Art. 12 (record-keeping) and Art. 14 (human oversight) compliance evidence bundle for an AI agent. Returns the cryptographic audit chain, Merkle epoch roots, delegation certificate, trust score, and violation history — sufficient for regulatory submission. The chain_verification.is_valid field proves the audit trail has not been tampered with.
verify_jurisdictionFreeVerify that an AI agent is authorized to operate in a specific H3 territorial cell. Checks the GNS-AIP delegation certificate: signature validity, temporal bounds, H3 cell authorization, and facet authorization. Returns a structured result indicating whether the agent may proceed.
Public scan report
scanner v0.1.9 · 2026-09-27 · same rubric, same numbers if you re-run it
- –Code scanremote-only server, no package to scann/a
- Live reliabilityremote reachable in 294ms20/20
- Tool poisoning3 tool descriptions checked15/15
- Auth qualityopen endpoint, read-only tools10/15
- Maintenancelast push 23 days ago15/15
- Maintainer identitynamespace and repository owner differ4/10
How Agentcore compares
Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.
- Safety score
- 85/100
- higher than 70% of 48,255 servers
- Tools
- 3
- the typical server has 7
- Tools that change things
- 0%
- typical: 0%. Fewer write tools means less an agent can break
- Uptime
- 100%
- of our recent checks; the typical remote server answers 100%
- Response time
- 165 ms
- faster than 70% of 30,031 servers
What its tools let an agent do: 2 read · 1 create
Install Agentcore in Claude Code, Cursor or VS Code
claude mcp add --transport http mcp-agentcore https://geiant-agentcore-production.up.railway.app/mcp
What the publisher says
From the Agentcore repository's README, as published. We do not edit it. Read it on GitHub
🐜 GEIANT
Geo-Identity Agent Navigation & Tasking
The first AI orchestrator that knows where it is.
What is GEIANT?
GEIANT is a geospatial AI agent orchestration runtime built on the GNS Protocol. It is a direct competitor to LangChain, CrewAI, and AutoGen — with one structural difference: every agent has a cryptographic identity bound to an H3 territory, and every task is routed through a compliance enforcement layer before dispatch.
LangChain doesn't know where it is. GEIANT does.
The Ant Metaphor
In GEIANT, agents are ants. Ants are the canonical model of emergent collective intelligence from simple agents following spatial rules. Ant Colony Optimization (ACO) is a real algorithm used in routing and pathfinding. The colony is the orchestrated workflow. The territory is the H3 grid.
- An ant (AntIdentity) is an agent with a GNS Ed25519 keypair + H3 territory + facet scope + trust tier
- The colony is a multi-agent workflow where ants delegate and collaborate
- The nest is the Agent Registry — the directory of available ants
- The trail is the Virtual Breadcrumb chain — the Proof-of-Jurisdiction audit log
Architecture
L6 Multi-Agent Orchestration — task graph, delegation chain enforcement
L5 IDUP Settlement Layer — Stellar payment routing per operation
L4 MCP GIS Switchboard — GDAL / PostGIS / ArcGIS / QGIS / EO models
L3 Perception Layer — IBM Prithvi, Microsoft Clay (EO Foundation Models)
L2 Geometry Validation Layer — GEOS-backed guardrails, self-intersection detection
L1 Geospatial Router — H3 jurisdiction matching, TierGate enforcement
L0 Agent Identity & Registry — Ed25519 + H3 + GNS-AIP (LIVE in gns-node)Phase 0 (this repo): L0 + L1 + L2 core logic, tests, and API scaffold.
Monorepo Structure
geiant/
├── packages/
│ ├── core/ @geiant/core — types, router, validation, registry, jurisdiction
│ └── sdk/ @geiant/sdk — developer-facing SDK (Phase 1)
├── apps/
│ └── api/ @geiant/api — Express REST API
└── docs/ Architecture docs, ADRsThe Four Router Gates
Every task passes through four gates before any ant is selected:
- Signature Verification — Ed25519 signature over canonical task JSON
- Jurisdiction Resolution — H3 cell → country → regulatory frameworks (GDPR, EU AI Act, FINMA, CCPA...)
- Delegation Chain Validation — human GNS identity → delegation cert → scope cells + facets + expiry
- Geometry Pre-flight — GEOS-backed validation: self-intersection, coordinate transposition, unclosed rings
Only tasks that pass all four are dispatched. The router is a compliance enforcement point, not a load balancer.
Getting Started
# Install dependencies
pnpm install
# Run tests
pnpm test
# Start dev API
pnpm dev
# Test routing endpoint
curl http://localhost:3100/health
curl http://localhost:3100/registry
curl http://localhost:3100/jurisdiction/8728d55ffffffff # Rome H3 cell
curl http://localhost:3100/route/testCompetitive Position
Patent & IP
GNS Protocol Provisional Patent #63/948,788 (Proof-of-Trajectory). GEIANT's Proof-of-Jurisdiction is a second patentable claim extending this.
License
Protocol spec: CC-BY Core library: Apache 2.0 Managed runtime: Proprietary
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
Agentcore: common questions
- Is Agentcore MCP server safe?
- Yes, by our scan: it is graded A (85/100). Read the Agentcore safety report
- How do I install Agentcore?
- It runs remotely at geiant-agentcore-production.up.railway.app. Add it to Claude Code, Cursor, VS Code or Claude Desktop with the snippets above, or call it through the mcp.market gateway without installing anything.
- Does Agentcore need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is Agentcore maintained?
- The last commit was 60 days ago (2026-09-05). The latest release is v0.1.0.
- Is Agentcore up?
- 100% of our last 28 checks got an answer. We check remote servers about four times a day.
- How does Agentcore compare with other MCP servers?
- safety score 85/100 (higher than 70% of 48,255 servers); tools 3 (the typical server has 7); tools that change things 0% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.