mcp-review
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| VERDICT_CLI | No | Absolute path to an apatureai/verdict build. Without it, findings come from a golden fixture and every payload is marked unjudged. | |
| MODEL_API_KEY | No | API key for MODEL_BASE_URL. Must be set together with MODEL_BASE_URL. | |
| MODEL_BASE_URL | No | OpenAI-compatible chat-completions endpoint verdict calls for real judgments. Must be set together with MODEL_API_KEY. | |
| BASTION_ALLOWED_HOSTS | No | Comma-separated hosts this server may review. Defaults to preview.example.com; every other host is rejected as DOMAIN_UNVERIFIED. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| design_reviewA | Submit an asynchronous, metered design review for a tenant-authorized HTTPS preview. This tool never edits code. Reuse client_request_id on retries, then poll design_review_get no faster than poll_after_ms. |
| design_review_getA | Get status or a compact, focused, or evidence view for an existing review job. Poll no faster than the returned poll_after_ms. Result reads do not consume review units. Act on review.measurements.violations unconditionally: they are computed from the captured DOM and are true whether or not a model ran. Act on review.findings only when provenance.model_backed is true and coverage.state is full or partial. |
| design_recheckA | Submit a metered recheck for findings from a completed review after the customer's agent changes the UI. This tool never edits code. Unchanged targets and exhausted recheck loops are rejected without running judgment. Check recheck.provenance before acting: when nothing judged the target every outcome is unjudged with a null confidence. Reuse client_request_id on retries. |
| design_review_cancelA | Request best-effort cancellation of a queued or running review job. Terminal jobs keep their existing state. Cancellation does not edit customer systems and consumes no review units. |
| design_review_panel_actionB | Route an interaction from the interactive review panel: return a grounded finding's fix for the coding agent to apply, or the refs to re-verify. This tool never edits code. An advisory finding returns human_only, never an automatic fix, and a review nothing judged returns unjudged with no fix at all. Reads only; consumes no review units. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool has a clearly distinct purpose: submit review, get status, recheck, cancel, and panel action. There is no overlap between submission and retrieval, and cancellation is separate from panel actions. The descriptions further differentiate them, making misselection unlikely.
All tools share the 'design_' prefix, but the suffix varies: 'review', 'review_get', 'recheck', 'review_cancel', 'review_panel_action'. While 'design_review' is a verb phrase, others combine 'review' with a noun (get, cancel) and 'recheck' is a different verb. The pattern is mostly consistent and readable, but not perfectly uniform.
Five tools is well within the recommended 3-15 range and matches the scope of a review service. Each tool covers a distinct operation in the review lifecycle without redundancy or bloat.
The surface covers the core lifecycle: submit, poll/retrieve, recheck, and cancel, plus an interactive panel action for post-review interactions. Minor gaps like listing all past reviews are not essential for the primary workflow, and the tool set is adequate for its stated purpose.