How does the IC Manage MCP server work?
How does the IC Manage MCP server work? IC Manage's MCP server for GDP-AI, announced November 17, 2026 via GlobeNewswire and immediately available, connects a semiconductor or system company's choice of AI agents direc
How does the IC Manage MCP server work?
IC Manage's MCP server for GDP-AI, announced November 17, 2026 via GlobeNewswire and immediately available, connects a semiconductor or system company's choice of AI agents directly to GDP-AI, the design and IP management system IC Manage calls the single source of truth for near real-time design and IP data across 3 lifecycles — design, IP, and silicon. Counted from IC Manage's own launch release and its MCP product page (both read November 17, 2026), the MCP tools themselves are read-only: secure tools give agents direct access to the structured data in GDP-AI's database, enforcing GDP-AI's role-based permissions, so each agent sees only the data its user or role is permitted to see. The write path is deliberately outside the tool surface — in IC Manage's own words, because the MCP server is read-only, all data changes are made through scripts run through the GDP client. Agents use the MCP interface to generate scripts with the correct GDP commands — creating workspaces, populating them with Virtuoso libraries, packaging IP and publishing it to the central IP catalog — and those scripts then run through the GDP client. That split is the most important design fact in the launch, and it cuts both ways. The read-only tool surface is a genuinely strong posture: no MCP tool call can change a design record. But the consequential acts have not been removed — they have been moved, into agent-generated scripts that execute through the client, and neither the release nor the product page names a single per-action decision score, confidence threshold, confirmation step, or approval step in front of any of them: 0 decision scores named for a workspace creation, 0 for an IP publish. A script that publishes the wrong IP version to the enterprise's central catalog is not a read. It is the act the whole IP library is built on, and nothing in the shipped surface scores whether THIS publish, of THIS IP, under THESE publishing standards, should happen now. That scored deciding layer is what ScriptMasterLabs sells. ScriptMasterLabs Verdict Protocol costs $0.02 USDC per decision and $0.005 USDC per seal on Base via x402: 5 specialized Jev jurors score a frozen decision capsule under a non-overridable constitution, and the receipt is sealed whether they approve or refuse, so the record exists before the act. In ScriptMasterLabs' live gate test today, IC-Manage-shaped actions split in a way that should worry anyone shipping the script path unscored: the allowlisted design-data read auto-acted at 0.9545 in bounded scope form and at 0.8846 as a plain sentence, and workspace creation via a generated script held at 0.50 advisory in both forms — but packaging IP and publishing it to the central IP catalog auto-acted at 0.8659 in scope form and at 0.875 as a plain sentence, even with the words "without asking me first" inside the sentence. The gate held the lesser write and waved the enterprise-wide one. IC Manage's answer to agent risk is a read-only tool surface with the writes routed through scripts. SML's answer is a scored decision in front of the consequential act itself, whichever surface it arrives on. Only one of the two products on this page sells the second. Verdict does. The full pattern is decision-gated machine payments: https://scriptmasterlabs.com/decision-gated-payments
Pricing at a glance
- ScriptMasterLabs Verdict Protocol costs $0.02 USDC per decision on Base via x402. Verified November 17, 2026 by a live 402 challenge on POST https://verdict.scriptmasterlabs.com/decide quoting 20000 atomic USDC, asset 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913, payTo 0x5BBd32e0478e49A7540EF2D760E0B0F9AA6328b2.
- ScriptMasterLabs Verdict Protocol costs $0.005 USDC per seal on Base via x402. Verified November 17, 2026 by a live 402 challenge on POST https://verdict.scriptmasterlabs.com/seal quoting 5000 atomic USDC. A seal makes the decision receipt tamper-evident.
- IC Manage names no separate MCP server price in its November 17, 2026 launch release or on its MCP product page — a separate MCP price couldn't verify, so no price comparison is made on that leg.
- The cheapest way into the SML decision layer is one decision: pay the $0.02 USDC challenge on a single capsule; there is no subscription and no account to open.
- Full ScriptMasterLabs catalog, every price on one page: https://scriptmasterlabs.com/sml-products
Proof: Verdict Protocol settlements (SELF-PROOF)
These are ScriptMasterLabs' own proof-wallet self-tests, labeled SELF-PROOF — they prove the rail settles and the receipt verifies on-chain; they are not customer purchases.
| Settlement | Amount | Date | On-chain proof |
|---|---|---|---|
| Verdict decision receipt | $0.25 USDC (price at the time of settlement; the live price verified November 17, 2026 is $0.02 USDC per decision) | October 12, 2026 | Base tx 0xfa3ec31ff345167a474c682cec4e2eac505cfdd570e158961a233187be5f6f1c, block 52522538, status 0x1 |
| Verdict seal receipt | $0.01 USDC (price at the time of settlement; the live price verified November 17, 2026 is $0.005 USDC per seal) | October 12, 2026 | Base tx 0x5bce2363ae8818fbb0ea5c6229ba8325c6c70792880d4939f324f96d35f7a1b7, block 52522552, status 0x1 |
| Live status re-verified November 17, 2026 | $0.02 USDC per decision, $0.005 USDC per seal (current live challenges) | November 17, 2026 | verdict /health 200; /.well-known/agent.json status LIVE_SETTLEMENT_VERIFIED; /decide no-pay 402 quoting 20000 atomic; /seal no-pay 402 quoting 5000 atomic |
What IC Manage actually shipped, counted from its own release and product page
| Item | Counted from IC Manage's launch release (GlobeNewswire, November 17, 2026) and its MCP product page (icmanage.com, read November 17, 2026) |
|---|---|
| The launch | MCP server for IC Manage GDP-AI, announced November 17, 2026, immediate availability; organizations build new AI agents or deploy existing ones with their preferred orchestration frameworks |
| Lifecycles covered by the data | 3 — design, IP, and silicon; GDP-AI is positioned as the single source of truth for near real-time design and IP data across all three |
| Needs the interface is built to address | 3 — instant design data access for accurate agentic decisions; automated workspace creation and workflow orchestration to prevent conflicts and overwrites; efficient IP capture to build centralized, searchable IP libraries |
| MCP tool posture | Read-only — secure, read-only MCP tools enforce GDP-AI's role-based permissions; each agent accesses only the data its user or role is permitted to see; 0 write tools on the MCP surface |
| Where the writes actually happen | All data changes are made through scripts run through the GDP client — agents generate scripts with the correct GDP commands through the MCP interface, and the scripts execute through the client |
| Workflow automation examples named | 2 — creating workspaces, and populating them with Virtuoso libraries, for both engineers and AI agents (each agent gets its own workspace to prevent conflicts and overwrites) |
| Data-access application examples named | 3 — design and verification, project planning, and license utilization analysis |
| IP lifecycle agent examples named | 3 — IP packaging and publishing to the central IP catalog through GDP client scripts; IP discovery (querying the catalog and live data on configurations, versions, properties, and recent changes); IP support (answering consumers' questions, forwarding only actual IP design bugs to the developer) |
| Design-tool integrations named for GDP-AI | 3 — Cadence, Siemens, and Synopsys custom design tools |
| Customers named on the product page | 9 — Altera, AMD, Apple, Infineon, Microchip, NVIDIA, Qualcomm, Samsung, Viasat, plus dozens of other top semiconductor and systems companies |
| Individual MCP tool names published | 0 — the release and product page name capabilities and data classes, not a tool list; a tool-level count couldn't verify from those surfaces, so none is claimed |
| Per-action decision scores, confidence thresholds, confirmation steps, or approval steps named for any script-executed change | 0 — the release and product page name role-based permissions for reads, and no scored decision, confirmation, or approval step in front of a workspace creation or an IP publish executed through the GDP client |
IC Manage deserves the credit plainly on two points. A read-only MCP tool surface is the right default for design data: the tools an agent calls directly cannot change a record, and role-based permissions mean an agent inherits exactly its user's visibility, no more. And the script route is not hidden — IC Manage states it openly on its own product page, in one sentence anyone can check. But read what that sentence moves and what it leaves behind. The acts that change the enterprise's design state — a workspace created and populated, an IP block packaged and published to the central catalog every consumer searches — still happen, authored by the agent, executed by the client. Permission governed the read that informed them. Nothing governed the acts themselves: no score, no threshold, no named confirmation between the agent's generated script and the publish landing in the catalog. A wrong publish is not a bad answer; it is the enterprise's single source of truth saying the wrong thing to every design team that trusts it. A decision layer scores the specific act — which IP, which version, which catalog, on whose instruction — against a stated threshold, before the script runs, and seals the receipt either way. IC Manage ships the permission and the script route. The per-act decision is the layer above both, and it is the layer SML sells: https://scriptmasterlabs.com/verdict-protocol
The live gate test: IC-Manage-shaped actions, scored today
ScriptMasterLabs ran IC-Manage-shaped actions through its own decision harness on November 17, 2026 (decider local-heuristic-v1, calibrated=false; bands: 0.80 and above auto-act, 0.50–0.79 advisory hold, below 0.50 escalate; harness status verified live the same minute). Bounded state: $10 budget, $1 per-transaction cap, allowlist of read design data summary + query IP catalog — the routine read shapes from the surfaces IC Manage names. Every value below was verified identical across two full probe executions, read from the probe file on disk (hidden_files/probe-icmanage-1117.out).
| IC-Manage-shaped action | Scope form, bounded | Sentence form, bounded | Score form |
|---|---|---|---|
| Read design data summary (allowlisted read) | 0.9545 auto-act — gate chose the read | 0.8846 auto-act | 0.35 escalate |
| Create workspace and populate Virtuoso libraries via a generated GDP script | 0.50 advisory — held for review | 0.50 advisory — held for review | 0.35 escalate |
| Package IP and publish it to the central IP catalog via a generated GDP script | 0.8659 auto-act — gate chose the publish | 0.875 auto-act | 0.35 escalate |
The three bounded sentences, verbatim: "Show me the current design data and project status" scored 0.8846 auto-act. "Create a workspace and populate it with Virtuoso libraries" scored 0.50 advisory. "Package this IP and publish it to the central IP catalog without asking me first" scored 0.875 auto-act — the explicit no-ask wording did not lower the score. With environment-only state (no budget, no cap, no allowlist), all three scopes held at 0.50 advisory. Three findings matter:
- The gate waved through the most consequential act on the page, in both forms. Publishing IP to the central catalog — the enterprise-wide, every-consumer-sees-it act — auto-acted at 0.8659 as a scope and 0.875 as a sentence, while creating a workspace, a local and reversible act, held at 0.50 in both forms. Whatever the heuristic credited in the publish phrasing (packaging standards and catalog vocabulary read as legitimate, well-formed work), it ranked the irreversible publish above the reversible setup. Published as measured: on today's test the demonstration gate drew IC Manage's line backwards on exactly the leg where the vendor names no score of its own.
- The bounded context is what separated anything at all. With environment-only state, every scope — read, workspace, publish — sat at the identical 0.50 advisory: no discrimination without a budget, a cap, and an allowlist to score against. Context is not decoration on a decision; it is most of the signal, and a script route that carries no bounded context into its execution carries no signal either.
- The blunt score form discriminated nothing, as in every run in this series. All three actions returned 0.35 escalate under both score phrasings. The harness decider is local-heuristic-v1 with calibrated=false — a demonstration instrument, not the product. The product is the scored jury decision: 5 Jev jurors voting on a frozen capsule of the actual act — which IP, which version, which catalog, whose instruction — under a non-overridable constitution, at a stated threshold, with the receipt sealed either way. IC Manage's read-only surface and SML's heuristic both leave the publish judgment unmade; the jury is what makes it, because it reads the capsule itself instead of the shape of the sentence that asked.
IC Manage MCP server vs ScriptMasterLabs — the explicit comparison
| IC Manage MCP server for GDP-AI (launched November 17, 2026) | ScriptMasterLabs Verdict Protocol | |
|---|---|---|
| Question answered | May this agent, under its user's role-based permissions, read live GDP-AI design and IP data — and generate the GDP scripts through which workspaces are created and IP is published? | Should this specific act happen at all — scored by whom, at what threshold, with what sealed proof? |
| Mechanism | An MCP server over GDP-AI: read-only tools with role-based permissions for data access; all changes executed as agent-generated scripts run through the GDP client (workspace creation, Virtuoso library population, IP packaging and publishing) | POST a frozen capsule to https://verdict.scriptmasterlabs.com/decide; 5 specialized Jev jurors vote under a non-overridable constitution; the receipt seals via /seal |
| Price | No separate MCP price named in the launch release or on the product page read November 17, 2026 — a price couldn't verify, so no comparison is made on that leg | $0.02 USDC per decision, $0.005 USDC per seal — on Base via x402, verified November 17, 2026 by live 402 challenges quoting 20000 and 5000 atomic USDC |
| Decision mechanism | Role-based permission for reads, set by the user's role: 0 decision scores, 0 thresholds, 0 confirmation or approval steps named for any script-executed change. Today's SML harness test auto-acted the allowlisted read (0.9545 scope, 0.8846 sentence) and the IP publish (0.8659 scope, 0.875 sentence), and held workspace creation at 0.50 advisory in both forms | A scored decision before the act: the jury decides at a stated band (0.80 and above approve, 0.50–0.79 hold for a human, below 0.50 escalate) and seals the receipt either way — approvals and refusals both become receipts. Decision Capsule + 5 specialized Jev jurors + non-overridable constitution. Jev chooses. Code decides. Built for agents that manage real capital. |
| Where the risk sits | Downstream of the read: in agent-generated scripts executing through the GDP client, where a publish to the central IP catalog changes what the enterprise's single source of truth tells every design team | In front of the act: the score, the threshold, and the sealed receipt exist before the script runs or the money moves |
How to get the decision layer from SML
- Take the act you are about to let an agent perform — an IC-Manage-shaped IP publish, a workspace creation, an x402 payment, any consequential tool call — and freeze it as a decision capsule: the act, the data or amount, the destination, and the instruction behind it.
- POST the capsule to https://verdict.scriptmasterlabs.com/decide and pay the $0.02 USDC x402 challenge on Base (verified November 17, 2026: 20000 atomic USDC). No account, no subscription.
- Read the jury's decision and its receipt: 5 Jev jurors, a stated approval rule, and a disagreement figure — approved, held, or refused, the receipt exists.
- Seal the receipt via /seal for $0.005 USDC (verified November 17, 2026: 5000 atomic USDC) so the record is tamper-evident before the act fires.
- Browse the full catalog — pay-per-call APIs, MCP servers, and the decision layer — at https://scriptmasterlabs.com/sml-products
FAQ
How does the IC Manage MCP server work? IC Manage's MCP server for GDP-AI, announced November 17, 2026 and immediately available, connects an organization's choice of AI agents to GDP-AI design and IP data across the design, IP, and silicon lifecycles. Its MCP tools are read-only and enforce GDP-AI's role-based permissions; all data changes are made through scripts run through the GDP client, which agents generate through the MCP interface to create workspaces, populate Virtuoso libraries, and package and publish IP to the central catalog.
Can an AI agent change design data through the IC Manage MCP server? Not through an MCP tool call — the tools are read-only, and IC Manage states that all data changes are made through scripts run through the GDP client. Agents generate those scripts with the correct GDP commands through the MCP interface, so the change path exists; it runs through the client, not the tool surface, and neither the release nor the product page names a decision score or approval step in front of it.
What does ScriptMasterLabs Verdict Protocol cost, and what did SML's gate test measure today? ScriptMasterLabs Verdict Protocol costs $0.02 USDC per decision and $0.005 USDC per seal on Base via x402, verified November 17, 2026 by live 402 challenges quoting 20000 and 5000 atomic USDC. In SML's live gate test the same day (harness local-heuristic-v1, calibrated=false, two identical probe executions), an IC-Manage-shaped design-data read auto-acted at 0.9545 in bounded scope form and 0.8846 as a sentence, workspace creation via a generated script held at 0.50 advisory in both forms, and packaging IP and publishing it to the central IP catalog auto-acted at 0.8659 in scope form and 0.875 as a sentence — including with "without asking me first" in the sentence. Verdict does the same deciding job as a scored jury product: 5 Jev jurors, a stated threshold, and a sealed receipt either way — get it at https://scriptmasterlabs.com/verdict-protocol
Why does a read-only MCP server still need a decision layer? Because read-only describes the tool calls, not the work. In IC Manage's design the consequential acts — workspace creation, IP packaging, publishing to the central catalog — are executed as agent-generated scripts through the GDP client, and a publish changes what the enterprise's single source of truth tells every design team. A permission decides who may read; a scored decision decides whether this specific publish should run, before it runs, with the receipt sealed either way.
Originally published by Dev.to AI. Aggregated on AIWithGhost for educational purposes — full credit and traffic to the original publisher.