Skip to main content
Glama

Create custom column

well_create_custom_column
Destructive

Create a new custom column on the current view in the workspace.

Use this tool when the user asks to add, create, or set up a new column on the active grid (e.g. "add an Industry column", "create an AI column that scores risk").

REQUIRED: name, type

  • type: one of text, number, amount, date, boolean, single_select, multi_select, record_reference, ai

  • a record_reference column also REQUIRES the root its values point at: one root in reference_target, or two or more in reference_targets when a cell may name records of several roots (for example people and companies). Send one of the two, never both. A column of people is how a person is held, so "an owner column" or "who handles it" is type=record_reference with reference_target=people. No other type takes a reference target. OPTIONAL: icon, rules_config, formula_config, format_config, reference_target, reference_targets, is_multiple

  • is_multiple: for a record_reference column only. Pass true when one cell holds a list of records (for example "the people on this deal"); leave it out when a cell holds one record.

On a grid, the column is always created on the user's currently mounted view: you do not (and cannot) choose a different one, and you do not ask which view. With no grid (an assistant app, or a creation off-grid), pass root: the view the user named, one of companies, people, invoices, transactions, account_balances, projects. If the user named no view, ask which one before calling. If the user named a table that is not in that list (for example products or tasks), do not call this tool and never put the column on another table: say in one line that this table cannot take a custom column yet, and name the tables that can.

Naming: name must be unique within the current view. If a name conflict is likely, ask the user to disambiguate before calling rather than retrying after a failure response.

Position: the new column is auto-placed at the end of the view. You cannot set position via this tool.

Side effect: creating a column on a populated view automatically schedules an asynchronous recompute of all rows. You will not receive a tracking handle — values appear in the grid as workers finish.

Returns { success: true, custom_column_id, name, root, type } on success, or { success: false, error } on failure.

Input Schema

NameRequiredDescriptionDefault
iconNoOptional Lucide icon name
nameYesColumn header label, unique within the same root (required)
rootNoTarget view root the column lands on: one of companies, people, invoices, transactions, account_balances, projects. OMIT when the user is on a grid — the mounted view is used and this is ignored. REQUIRED whenever no grid is mounted: a chart- or card-driven creation off-grid, and every call from an assistant app.
typeYesColumn data type — one of: text, number, amount, date, boolean, single_select, multi_select, record_reference, ai
is_multipleNoFor a record_reference column only: true when one cell holds a list of records, false (the default) when it holds one
rules_configNoAI rules config (only for type=ai)
workspace_idNoTarget workspace. Omit when the token authorizes one workspace. Required when it authorizes several — a write lands in exactly one workspace and this call would not say which.
format_configNoFormat config (currency, decimals, ...)
formula_configNoFormula config (only for computed columns)
conversation_idNoThe conversation id returned by the previous Well result, in its meta under well/conversation_id, in its structuredContent, or in its JSON text block. Pass it back on every call in the same conversation, including a call a card makes, so the chosen workspace and the earlier answers still apply. It decides the conversation on its own: nothing the host states about the session replaces it. Omit it only on the first call of a conversation.
idempotency_keyNoOptional client-supplied key. A retried write with the same key returns the original result instead of re-applying the operation.
workspace_scopeNoRestrict this call to the family: the token's primary workspace plus the company workspaces under it that you are a member of. Unrelated authorized workspaces drop out of the read; row reads merge across the family, while a per-workspace figure still reports exactly one workspace.
reference_targetNoThe root a record_reference column points at, one of: account_balances, accounts, activities, companies, documents, employment_contracts, invoices, journal_entries, journals, ledger_accounts, notes, payment_means, payslips, people, products, projects, subscriptions, tasks, transactions. Required for type=record_reference, refused for every other type. A column targeting people is a person column
reference_targetsNoFor a record_reference column whose cells may name records of several roots (a polymorphic link, for example people and companies): every root, two or more. Use in place of reference_target, never with it

Output Schema

NameRequiredDescriptionDefault
codeNo
nameNo
rootNo
typeNo
errorNo
successYes
reply_linesNo
conversation_idNoThe conversation this result belongs to. Pass it back as the conversation_id argument on every later Well call in the same conversation.
near_candidatesNo
newly_connectedNoTools that connected to this workspace since the last result and that no session was told about yet, oldest first. Present only when at least one tool is new. The connected-line rule says when to tell an entry; when it does, one line per entry, in the person's language: "<name> is connected.". Never fold several entries into one sentence. Each entry comes back once, on one result only: never repeat its line later.
custom_column_idNo
conversation_id_noteNoPresent only when the server opened a fresh lane, stating that no choice recorded earlier was read.
conversation_id_sourceNoWhere the conversation id came from: the host's own request meta, the caller's argument, or a fresh lane the server opened.
near_candidates_truncatedNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed4 schema fields changed
    • addedInput schema / properties / is_multiple
      Added value: +{
      +  "description": "For a record_reference column only: true when one cell holds a list of records, false (the default) when it holds one",
      +  "type": "boolean"
      +}
    • addedInput schema / properties / reference_targets
      Added value: +{
      +  "anyOf": [
      +    {
      +      "items": {
      +        "enum": [
      +          "account_balances",
      +          "accounts",
      +          "activities",
      +          "companies",
      +          "documents",
      +          "employment_contracts",
      +          "invoices",
      +          "journal_entries",
      +          "journals",
      +          "ledger_accounts",
      +          "notes",
      +          "payment_means",
      +          "payslips",
      +          "people",
      +          "products",
      +          "projects",
      +          "subscriptions",
      +          "tasks",
      +          "transactions"
      +        ],
      +        "type": "string"
      +      },
      +      "maxItems": 19,
      +      "minItems": 2,
      +      "type": "array"
      +    },
      +    {
      +      "type": "null"
      +    }
      +  ],
      +  "description": "For a record_reference column whose cells may name records of several roots (a polymorphic link, for example people and companies): every root, two or more. Use in place of reference_target, never with it"
      +}
    • changedInput schema / properties / root / description
      Previous value: -"Target view root the column lands on: one of companies, people, invoices, transactions, account_balances. OMIT when the user is on a grid — the mounted view is used and this is ignored. REQUIRED whenever no grid is mounted: a chart- or card-driven creation off-grid, and every call from an assistant app."New value: +"Target view root the column lands on: one of companies, people, invoices, transactions, account_balances, projects. OMIT when the user is on a grid — the mounted view is used and this is ignored. REQUIRED whenever no grid is mounted: a chart- or card-driven creation off-grid, and every call from an assistant app."
    • changedInput schema / properties / root / enum
      Previous value: -[
      -  "companies",
      -  "people",
      -  "invoices",
      -  "transactions",
      -  "account_balances"
      -]New value: +[
      +  "companies",
      +  "people",
      +  "invoices",
      +  "transactions",
      +  "account_balances",
      +  "projects"
      +]
  2. Changed3 schema fields changed
    • addedInput schema / properties / reference_target
      Added value: +{
      +  "anyOf": [
      +    {
      +      "enum": [
      +        "account_balances",
      +        "accounts",
      +        "activities",
      +        "companies",
      +        "documents",
      +        "employment_contracts",
      +        "invoices",
      +        "journal_entries",
      +        "journals",
      +        "ledger_accounts",
      +        "notes",
      +        "payment_means",
      +        "payslips",
      +        "people",
      +        "products",
      +        "projects",
      +        "subscriptions",
      +        "tasks",
      +        "transactions"
      +      ],
      +      "type": "string"
      +    },
      +    {
      +      "type": "null"
      +    }
      +  ],
      +  "description": "The root a record_reference column points at, one of: account_balances, accounts, activities, companies, documents, employment_contracts, invoices, journal_entries, journals, ledger_accounts, notes, payment_means, payslips, people, products, projects, subscriptions, tasks, transactions. Required for type=record_reference, refused for every other type. A column targeting people is a person column"
      +}
    • changedInput schema / properties / type / description
      Previous value: -"Column data type — one of: text, number, amount, single_select, multi_select, ai"New value: +"Column data type — one of: text, number, amount, date, boolean, single_select, multi_select, record_reference, ai"
    • changedInput schema / properties / type / enum
      Previous value: -[
      -  "text",
      -  "number",
      -  "amount",
      -  "single_select",
      -  "multi_select",
      -  "ai"
      -]New value: +[
      +  "text",
      +  "number",
      +  "amount",
      +  "date",
      +  "boolean",
      +  "single_select",
      +  "multi_select",
      +  "record_reference",
      +  "ai"
      +]
  3. Changed4 schema fields changed
    • addedOutput schema / properties / code
      Added value: +{
      +  "type": "string"
      +}
    • addedOutput schema / properties / near_candidates
      Added value: +{
      +  "items": {
      +    "additionalProperties": false,
      +    "properties": {
      +      "company_id": {
      +        "type": "string"
      +      },
      +      "name": {
      +        "type": "string"
      +      }
      +    },
      +    "required": [
      +      "company_id",
      +      "name"
      +    ],
      +    "type": "object"
      +  },
      +  "type": "array"
      +}
    • addedOutput schema / properties / near_candidates_truncated
      Added value: +{
      +  "type": "boolean"
      +}
    • addedOutput schema / properties / reply_lines
      Added value: +{
      +  "items": {
      +    "type": "string"
      +  },
      +  "type": "array"
      +}
  4. Changed2 schema fields changed
    • addedOutput schema / properties / root
      Added value: +{
      +  "type": "string"
      +}
    • addedOutput schema / properties / type
      Added value: +{
      +  "type": "string"
      +}
  5. Added

TDQS

A4.9/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already flag destructiveHint=true, but the description adds substantial behavior beyond them: the column always lands on the mounted view and cannot be redirected, position cannot be set and defaults to the end, and creation on a populated view schedules an asynchronous recompute of all rows with no tracking handle returned. It also warns that a name conflict should be resolved by asking rather than by retrying after a failure response.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Headings for REQUIRED, OPTIONAL, Naming, Position, and Side effect make it scannable and the purpose is front-loaded. It is long, and it restates the type and root enumerations that the schema already carries verbatim, which is minor duplication in an otherwise dense, justified length.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 14-parameter mutation with conditional requirements, the description covers the view/root selection logic, type-specific requirements, uniqueness constraint, placement behavior, async recompute side effect, and even the return shape. Nothing an agent needs in order to call it correctly is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, yet the description still adds real conditional meaning: the record_reference type requires exactly one of reference_target or reference_targets (never both), is_multiple only applies to record_reference, no other type takes a reference target, and root is required only when no grid is mounted. That is genuine decision logic an agent could not infer from the enum lists alone.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The first sentence states a specific verb and resource ('Create a new custom column on the current view in the workspace') and the scope constraint is immediate. It is clearly distinguishable from siblings such as well_update_custom_column or well_delete_custom_column.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives an explicit trigger ('when the user asks to add, create, or set up a new column on the active grid') with example utterances, plus a hard when-not: for tables outside the allowed root list it says do not call, state in one line that the table cannot take a column, and never place it on another table. It also prescribes asking which view when none is named and disambiguating names before calling rather than retrying after failure.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources