Spreadshet Audit MCP
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
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 |
|---|---|
| audit_workbookA | Audits an .xlsx workbook: error values after an independent recalculation, typed-over formulas, formulas that break the pattern of their neighbours, hard-coded numbers inside formulas, approximate-match lookups, volatile functions, external links, hidden sheets/rows/columns and broken named ranges. Heuristic: it points at where models usually go wrong, it cannot prove a model is right. |
| inspect_cellA | Returns the formula (or value), the stored result if the file has one, and the cells the formula reads. Use it to follow a suspicious number back to its inputs. |
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 2 tools
The two tools are cleanly separated: audit_workbook performs a whole-file heuristic sweep, while inspect_cell zooms in on a single cell's formula, value, and precedents. The drill-down workflow is explicitly described, so an agent can always tell which to reach for.
Both names follow the same verb_noun snake_case convention (audit_workbook, inspect_cell), with verbs that accurately mirror each tool's scope. No stylistic deviations.
Two tools is at the thin edge for a spreadsheet auditing server; the surface is coherent but leaves no room for reporting, filtering, or remediation operations. It works as a minimal diagnose-then-drill pair, nothing more.
Detection and follow-up tracing are covered, but there is no way to export or summarize findings, fix/annotate flagged issues, or compare across workbooks. For a read-only audit domain this is workable, but agents will hit dead ends on any remediation or reporting task.