scopeguard-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| SCOPE | No | Path to a JSON scope document (CLI flag --scope PATH). Defines the declarative allow/deny rules (`name`, `description`, `allow[]`, `deny[]`, each rule being {"type": "host|url|cidr", "value": "..."}). | loopback-only built-in scope |
| LEDGER | No | Path to the append-only JSONL findings ledger file (CLI flag --ledger PATH). | ./scopeguard-findings.jsonl |
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 | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| check_targetA | Check whether a URL or host is inside the authorized testing scope. Returns an allow/deny verdict with the matching rule. Performs no network activity: no DNS lookups, no connections. |
| cvss_baseA | Compute the CVSS v3.1 Base score and qualitative severity for a vector string such as CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H. |
| redactA | Strip secrets, tokens and PII from text before it is pasted into a finding, a prompt or a report. Returns the cleaned text and counts. |
| add_findingA | Append a finding to the local ledger. The target is re-checked against the authorized scope and evidence is redacted automatically. Fails if the target is out of scope. |
| list_findingsC | Read findings back from the local ledger, newest last. |
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 distinct primary action: scope checking, CVSS scoring, redaction, reading findings, and appending findings. There is mild conceptual overlap because add_finding internally re-checks scope and redacts evidence, but the descriptions make clear that check_target and redact are standalone utilities while add_finding is the enforced write path.
Three tools follow a clear verb_noun snake_case pattern (check_target, list_findings, add_finding), but cvss_base is a noun phrase and redact is a bare verb. The set remains readable and all lowercase snake_case, but the convention is not uniform.
Five tools is well-scoped for a narrow scope-guard and findings-ledger server. Each tool maps to a necessary step in the workflow: validate scope, score severity, sanitize text, and read/write findings.
The surface covers scope checking, scoring, redaction, and append/read of findings, but the ledger lacks get-by-id, update, delete, or filtering operations, and there is no tool for managing or listing the authorized scope rules themselves. Some gaps may be intentional for an append-only ledger, but agents will hit dead ends when correcting or querying findings.