Format/lint JS, Python, HubL for automation platforms. Tools: format_code, format_json, ask_codefmt
Scored across 3 tools
ask_codefmt (Q&A), format_code (JS/Python/HubL), and format_json (JSON) each target a fairly distinct purpose. The only mild overlap is that JSON is arguably a form of code, so an agent might occasionally hesitate between format_code and format_json.
format_code and format_json follow a clean verb_noun snake_case pattern. ask_codefmt deviates by appending the brand name rather than an object noun, but it remains readable and consistent in verb style.
Three tools is on the thin side but reasonable for a focused formatting/linting service. Each tool earns its place, though a slightly richer surface might be expected given the multi-platform scope described.
Covers the core needs: code formatting/linting, JSON formatting (with minify and key sorting), and grounded Q&A. Minor gaps exist (e.g., no explicit supported-languages/platforms reference or config/option discovery tool), but core workflows are reachable.
Ask a natural-language question about codefmt: formatting and linting JavaScript, Python, and HubL for automation platforms (Zapier, n8n, Pipedream, Make, HubSpot). Returns a grounded answer drawn from codefmt's own content with cited source URLs and a confidence score (0-1, how much of the question the answer covers, weighted toward its rarer terms); off-topic or too-weakly-matched questions are refused rather than guessed.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | The natural-language question about codefmt. |
| Name | Required | Description |
|---|---|---|
| answer | Yes | The grounded answer, or the refusal text. |
| results | Yes | Ranked matching entries. |
| sources | Yes | Cited source URLs. |
| confidence | No | How much of the question the answer covers, weighted toward its rarer terms. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the readOnlyHint annotation, the description discloses return behavior (grounded answer with cited source URLs), the meaning and scale of the confidence score (0-1, weighting toward rarer terms), and the refusal policy for weak matches. That is exactly the behavioral context an agent needs and cannot get from the annotation or schema.
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?
It is a single dense paragraph with the core purpose front-loaded, followed by domain scope, return behavior, and refusal policy. Every sentence carries information, though the heavy parenthetical listing makes it slightly more crowded than necessary.
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 one-parameter Q&A tool, the description covers domain, expected input, return shape, confidence semantics, and failure behavior. Even though an output schema exists (so return values needn't be explained), the added explanation of the confidence score is genuinely helpful and nothing material is missing.
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?
There is a single required parameter with 100% schema description coverage, so the schema already documents "query" fully. The description reinforces that the query must be natural-language and about codefmt, but adds no format, length, or syntax detail beyond the schema. Baseline 3 is appropriate.
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 states a specific verb and resource ("Ask a natural-language question about codefmt") and scopes the domain precisely (formatting/linting JS, Python, HubL for Zapier, n8n, Pipedream, Make, HubSpot). It is clearly distinct from format_code and format_json by the verb alone, but it never names or contrasts those siblings explicitly.
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?
It gives clear context for when the tool applies (NL questions about codefmt's subject matter) and a when-not rule ("off-topic or too-weakly-matched questions are refused rather than guessed"). It does not, however, tell the agent when to prefer this over the sibling formatting tools.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Format and lint JavaScript, Python, or HubL code for Zapier, n8n, Pipedream, Make, or HubSpot. Wraps the source in each platform's runtime shell before formatting so top-level await, bare return, and injected globals don't break the formatter. Returns formatted code plus lint diagnostics.
| Name | Required | Description | Default |
|---|---|---|---|
| code | Yes | Source code to format (≤50KB). | |
| language | No | Defaults to javascript. zapier, n8n, pipedream, make, and hubspot support python; hubl does not. | |
| platform | Yes | Target automation platform. |
| Name | Required | Description |
|---|---|---|
| error | Yes | Why the format failed, or null on success. |
| diagnostics | Yes | Lint findings with line and column. |
| formatted_code | No | The formatted source. Absent when the input was refused. |
| diagnostics_frame | No | Whether diagnostic positions refer to the submitted code or the formatted code. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With readOnlyHint=true the safety profile is already declared, yet the description adds real behavioral detail: it wraps source in each platform's runtime shell so top-level await, bare return, and injected globals don't break the formatter. That explains non-obvious internal behavior beyond the annotations. The 'returns formatted code plus lint diagnostics' sentence is partly redundant with the output schema.
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, zero filler. The primary purpose and supported languages/platforms are front-loaded, with the non-obvious wrapping rationale following. Every clause 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 three-parameter, enum-driven tool with an output schema covering the return shape, the description supplies purpose, platform scope, and the key formatting caveat. It is essentially complete, with minor room for stating the size limit or error behavior, both of which the schema already handles.
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 100%, including enum members and the python-support caveat for hubl, so the schema carries the parameter semantics. The description adds nothing to the code/language/platform parameters beyond repeating the platform names. Baseline 3 applies.
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?
States a specific verb pair (format and lint) with a specific resource (JavaScript, Python, or HubL code) and enumerates the target platforms. An agent can distinguish this from format_json (JSON) and ask_codefmt without opening any schema.
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 platform enumeration implies the usage context, but there is no explicit when-to-use or when-not-to-use statement and no routing to the siblings ask_codefmt or format_json. The agent must infer that non-JS/Python/HubL content belongs elsewhere.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Format (pretty-print or minify) a JSON document, optionally sorting object keys. Number literals are preserved exactly (64-bit IDs safe); parse errors report the exact line and column.
| Name | Required | Description | Default |
|---|---|---|---|
| json | Yes | The JSON text to format. | |
| indent | No | Indentation: '2' or '4' spaces, 'tab', or 'minify'. Defaults to '2'. | |
| sort_keys | No | Sort object keys alphabetically. Defaults to false. |
| Name | Required | Description |
|---|---|---|
| error | Yes | "line L, col C: message" on a parse error, or null. |
| formatted_json | No | The formatted document. Absent when the input was refused. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
readOnlyHint=true already declares a safe read, but the description adds genuine value beyond annotations: lossless preservation of number literals (64-bit IDs safe) and precise line/column error reporting. These are non-obvious behavioral traits that materially affect how an agent trusts and calls the 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 dense sentences, front-loaded with the core operation and then the two edge-case guarantees. No filler; every clause carries distinct information.
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?
With an output schema present and 100% parameter coverage, the description need not explain returns. It covers the operation, options, numeric fidelity, and error reporting, so it is nearly complete; the only real gap is the missing sibling/usage guidance.
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 100%, so the schema already documents json, indent, and sort_keys fully. The description echoes pretty-print/minify and optional key sorting but adds no syntax or format detail beyond the schema, which is the expected baseline.
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?
States a specific verb (format/pretty-print/minify) and resource (JSON document), so an agent immediately knows the operation. It does not, however, name or distinguish itself from the siblings format_code or ask_codefmt, leaving the JSON-vs-other-language split implicit.
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?
There is no explicit when-to-use, when-not-to-use, or alternative routing despite two sibling tools (format_code, ask_codefmt) that plausibly overlap. The agent must infer that this is the JSON-specific formatter on its own.
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.
ask_codefmt2 fields changedOutput schema / properties / results / items / propertiesAdded value: +{
+ "label": {
+ "type": "string"
+ },
+ "score": {
+ "type": "number"
+ },
+ "source": {
+ "type": "string"
+ }
+}Output schema / properties / results / items / requiredAdded value: +[
+ "label",
+ "source",
+ "score"
+]ask_codefmt1 field changedOutput schema / (root)Previous value: -nullNew value: +{
+ "properties": {
+ "answer": {
+ "description": "The grounded answer, or the refusal text.",
+ "type": "string"
+ },
+ "confidence": {
+ "description": "How much of the question the answer covers, weighted toward its rarer terms.",
+ "maximum": 1,
+ "minimum": 0,
+ "type": "number"
+ },
+ "results": {
+ "description": "Ranked matching entries.",
+ "items": {
+ "type": "object"
+ },
+ "type": "array"
+ },
+ "sources": {
+ "description": "Cited source URLs.",
+ "items": {
+ "type": "string"
+ },
+ "type": "array"
+ }
+ },
+ "required": [
+ "answer",
+ "sources",
+ "results"
+ ],
+ "type": "object"
+}format_code1 field changedOutput schema / (root)Previous value: -nullNew value: +{
+ "properties": {
+ "diagnostics": {
+ "description": "Lint findings with line and column.",
+ "items": {
+ "type": "object"
+ },
+ "type": "array"
+ },
+ "diagnostics_frame": {
+ "description": "Whether diagnostic positions refer to the submitted code or the formatted code.",
+ "enum": [
+ "input",
+ "output"
+ ],
+ "type": "string"
+ },
+ "error": {
+ "description": "Why the format failed, or null on success.",
+ "type": [
+ "string",
+ "null"
+ ]
+ },
+ "formatted_code": {
+ "description": "The formatted source. Absent when the input was refused.",
+ "type": "string"
+ }
+ },
+ "required": [
+ "diagnostics",
+ "error"
+ ],
+ "type": "object"
+}format_json1 field changedOutput schema / (root)Previous value: -nullNew value: +{
+ "properties": {
+ "error": {
+ "description": "\"line L, col C: message\" on a parse error, or null.",
+ "type": [
+ "string",
+ "null"
+ ]
+ },
+ "formatted_json": {
+ "description": "The formatted document. Absent when the input was refused.",
+ "type": "string"
+ }
+ },
+ "required": [
+ "error"
+ ],
+ "type": "object"
+}ask_codefmtformat_codeformat_jsonValidate and convert JSONL fine-tuning data across 11 AI providers. 13 tools.
Marketo MCP server for AI. 130 tools to operate Marketo from Claude, Cursor, or ChatGPT.
mtdt.io MCP: Salesforce DevOps platform connector (metadata deployments, backups and more)
Serves your design system and coding standards to coding agents, so they stop guessing.
Add one secure layer between your agents and this server.
