tablekit-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": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| table_convertA | Convert tabular data between formats, without losing or reshuffling rows. Use this whenever a table needs to change shape: CSV to Markdown, Markdown to JSON, HTML table to CSV, and so on. Rows are never silently dropped, and ragged rows are padded rather than truncated. Markdown output is column-aligned with East-Asian wide characters counted as two cells, so Chinese tables line up correctly. Args: text: The whole table as text. from_format: Source format. markdown, csv, tsv, json or html (aliases: md, htm). to_format: Target format. markdown, csv, tsv, json or html (aliases: md, htm). aligns: Markdown output only. Per-column alignment as a comma-separated list, e.g. "left,right,center", or one value applied to every column, e.g. "center". Empty means leave the default alignment. indent: JSON output indentation width. Default 2. |
| table_alignA | Re-align a messy Markdown table so every column lines up. Use this when a Markdown table is ragged, when pipes are not aligned in the source, when a row has the wrong number of cells, or when Chinese characters make the columns drift. This is a focused shortcut for table_convert(..., to_format="markdown"); it always outputs Markdown. Args: text: The table as text. Defaults to reading it as Markdown. from_format: How to read the input. markdown, csv, tsv, json or html. aligns: Per-column alignment, e.g. "left,right", or a single value. Empty means no explicit alignment markers. |
| table_transformA | Reshape a table: transpose, pick or rename columns, sort, dedupe, filter. Prefer this over editing rows by hand when the change is mechanical, because the whole table is processed at once and row order is preserved. Args: text: The whole table as text. operation: What to do. transpose | select_columns | rename_columns | sort | dedupe | drop_empty_rows | filter. from_format: How to read the input. markdown, csv, tsv, json or html. to_format: How to write the result. markdown, csv, tsv, json or html. columns: For select_columns: the column names to keep, in output order. A 1-based column number is also accepted, e.g. "2". renames: For rename_columns: a mapping of old name to new name. sort_by: For sort: the column to sort on. descending: For sort: set true for descending order. numeric: For sort: set true to sort numerically instead of as text. column: For filter: the column to test. operator: For filter: == | != | > | >= | < | <= | contains | startswith | endswith | empty | not_empty. value: For filter: the value to compare against. aligns: Markdown output only. Per-column alignment or a single value. |
| table_inspectA | Report the structure and data quality of a table before working on it. Returns row and column counts, a per-column summary (inferred type, empty count, distinct values, most frequent values) and a list of problems such as duplicate headers, ragged rows, fully duplicated rows and empty rows. Call this first when a table looks suspicious, then fix it with table_transform. Args: text: The whole table as text. from_format: How to read the input. markdown, csv, tsv, json or html. |
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 4 tools
The four tools have largely distinct purposes: align for realignment, convert for format changes, transform for reshaping, and inspect for analysis. However, table_align is a subset of table_convert (always outputs markdown) which could cause confusion when deciding which tool to use. This minor overlap prevents a perfect score.
All tool names follow the consistent pattern 'table_<verb>' using snake_case: table_align, table_convert, table_transform, table_inspect. The verbs are clear and predictable, making the API easy to navigate.
With 4 tools, the server covers the core operations (convert, transform, inspect, align) without bloat. While slightly on the lean side, the scope is focused and each tool justifies its existence. A couple more tools (e.g., merge or split) might be expected but the current count is reasonable.
The toolkit provides solid coverage for tabular data manipulation: format conversion, structural transformation, inspection, and alignment. Missing operations like merging tables or creating from scratch are not critical for the stated purpose, but there are minor gaps (e.g., no direct row insertion/deletion) that agents can work around using transform operations.