Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Jev Checkpoint Tests fail after dependency update. Choose: inspect dependency, trace code, or ask user."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Jev Checkpoint is a local MCP server that turns a bounded routing decision into one TypeSafe Jev choice question. It returns the chosen route, probabilities, confidence, and a thresholded recommendation. It never performs the selected action.
Use it when an agent has already gathered evidence and needs to choose among a small set of next actions, for example:
proceed, inspect callers, or ask the user;
routine review, deeper review, or human review; or
code, configuration, dependency, or insufficient evidence.
It is not a planner, reviewer, implementation model, or approval mechanism.
Related MCP server: genpark-edge-geo-distributed-request-routing-governor-skill
Install and build:
npm install
npm run buildSet your TypeSafe key:
export TYPESAFE_API_KEY="your-key"You can get the API key from e.g. Openrouter : https://openrouter.ai/typesafe/jev-1.13
Add this server to Claude Code's MCP settings (replace the path):
{
"mcpServers": {
"jev-checkpoint": {
"command": "node",
"args": ["/absolute/path/to/jev-checkpoint/dist/index.js"],
"env": { "TYPESAFE_API_KEY": "your-key" }
}
}
}Add this to your project's CLAUDE.md (or its equivalent agent
instructions):
When a bounded routing decision is uncertain and has explicit choices, call
`classify_decision` before proceeding. Include the decision, compact evidence,
allowed choices, and error cost. Treat its result as advisory. Never use it to
auto-commit, push, deploy, make clinical decisions, or send sensitive data to a
third party without approval.Claude Code then knows when to call the tool.
The optional reusable skill is .claude/skills/confidence-gate/SKILL.md.
classify_decision accepts:
{
"decision": "Choose the next safe action.",
"state": "Tests fail only after package X changed from 1.2 to 1.3.",
"choices": [
{ "id": "inspect_dependency", "description": "Inspect the package change and callers." },
{ "id": "inspect_code", "description": "Trace the failing code path." },
{ "id": "ask_user", "description": "Ask for missing reproduction details." }
],
"errorCost": "medium",
"threshold": 0.95
}The result is advisory. A confidence below threshold always returns
needs_deeper_review.
flowchart LR
A[Agent reaches an uncertain next step] --> B[Collect compact evidence]
B --> C{2-10 explicit routes?}
C -- no --> D[Use normal planning or investigation]
C -- yes --> E[classify_decision]
E --> F{Confidence meets threshold?}
F -- yes --> G[Take selected safe route]
F -- no --> H[Inspect more, use deeper model, or ask user]flowchart LR
A[Shadow mode: Jev recommendation] --> B[Normal workflow still runs]
B --> C[Record final route, findings, time, and cost]
C --> D{False-proceed rate and cost improve?}
D -- yes --> E[Allow the gate to suppress selected deeper calls]
D -- no --> F[Keep advisory-only or remove the gate]This server is advisory-only. It never performs the selected action, and a calling agent must preserve its own approval and safety rules.
This project makes no claims yet about cost, speed, or quality improvement. Those claims require a published benchmark against the intended workflow.
Do not send secrets, source code you are not allowed to disclose, or PHI to the TypeSafe API. A local MCP process does not make a hosted API local.
Return an advisory, confidence-gated route for a bounded decision. This tool never executes the route.
| Name | Required | Description | Default |
|---|---|---|---|
| state | Yes | Compact evidence available before the next action. | |
| choices | Yes | ||
| decision | Yes | The decision to make. | |
| errorCost | Yes | Impact if the selected route is wrong. | |
| threshold | No | Minimum confidence required for a recommendation. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full behavioral burden and does disclose one significant trait: 'This tool never executes the route.' Yet it leaves the confidence-gating behavior unexplained—particularly what happens when confidence falls below the threshold—and omits side effects and response semantics, which matter for an advisory tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two sentences with zero filler: the core purpose is front-loaded, and the critical behavioral caveat follows immediately. Every word earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 5-parameter tool with no output schema and no annotations, the description leaves key runtime behavior unexplained—most notably what happens when confidence is below the default 0.95 threshold and what the returned advisory route actually looks like. An agent would have to guess at the return shape and fallback behavior.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 80%, so the schema already documents most parameters, giving a baseline of 3. The description adds no per-parameter value and does not connect concepts like 'threshold' and 'errorCost' to the 'confidence-gated' behavior described.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description uses a specific verb ('Return') and a specific resource ('advisory, confidence-gated route for a bounded decision'), making the tool's advisory, non-executing purpose clear. With no sibling tools to differentiate from, it cannot distinguish itself from alternatives, but the core meaning is unambiguous.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The phrase 'for a bounded decision' provides implied usage context, and 'never executes the route' hints it is for advice only. However, there is no explicit when-to-use guidance, no prerequisites, and no exclusions—though the absence of sibling tools lowers the need for alternative-routing.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.1.0classify_decisionScored across 1 tool
With only a single tool, there is no possibility of confusion between tools. The tool's purpose is clearly described and distinct from any other potential operation.
The tool name follows a clean verb_noun pattern (classify_decision). Even though there is only one tool, the name is conventional and consistent with common MCP naming styles.
A single tool feels thin for most servers, and this one is no exception. However, the narrow scope of 'advisory classification only' makes the low count somewhat reasonable, but it remains on the borderline.
The tool fully covers its stated purpose of returning an advisory, confidence-gated route for a bounded decision. The explicit note that it never executes the route defines the boundary clearly, so there are no obvious missing operations within that scope.
The system of record for AI agent authority: playbooks, routed policy questions, reusable rules.
Deterministic contextual decision arbitration and action routing for autonomous software. Takes current state, context, or intent plus caller-supplied candidate actions, state transitions, routes, refusals, escalations, tools, or models and returns a deterministic ordered candidate field. Also provides persistent machine representations for memory, retrieval, indexing, and downstream coherence measurement.
Human-in-the-loop review and approval for AI agents. Audit trail, approval policies, native MCP.
Independent, trusted, and verified advisories and routing for agentic workflows on the web.
