Skip to main content
Glama
cubinet-code

paperless-ngx-mcp

by cubinet-code
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

create_workflow

Create a complete Paperless automation workflow in one call by defining triggers and ordered actions, such as stripping PDF passwords or tagging documents.

Instructions

Create a complete workflow — the automation Paperless actually runs — with its triggers and actions in one call. The workflow runs its actions in order whenever ANY of its triggers matches. Example: strip PDF passwords from API uploads = triggers [{type: 2, sources: [2]}] + actions [{type: 5, passwords: ["secret"]}]. Triggers and actions are given as full objects (same fields as create_workflow_trigger / create_workflow_action). Paperless rejects a remote OCR action (7) without a consumption-started trigger (1), and an apply-AI-suggestions action (8) with only consumption-started triggers.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
nameYes
orderNoRun order relative to other workflows (lower runs first)
actionsYesNew actions (any id is ignored)
enabledNoDefault true
triggersYesNew triggers (any id is ignored)

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed15 schema fields changedv3.2.1
    • changedInput schema / properties / triggers / items / properties / filter_custom_field_query / description
      Previous value: -"Custom-field filter as a JSON expression passed as a string. A condition is [field, operator, value], where field is the custom field's name or ID. Combine conditions with [\"AND\", [cond, …]] or [\"OR\", [cond, …]], and negate with [\"NOT\", cond]. Valid operators depend on the field's data type (e.g. exact, in, isnull, exists, icontains, gt, gte, lt, lte, range); pick an invalid one and Paperless names the valid ones. Examples: '[\"Amount\", \"gte\", 100]', '[\"AND\", [[\"Due\", \"lt\", \"2026-10-01\"], [\"Paid\", \"exact\", false]]]'."New value: +"Custom-field filter as a JSON expression passed as a string. A condition is [field, operator, value], where field is the custom field's name or ID. Combine conditions with [\"AND\", [cond, …]] or [\"OR\", [cond, …]], and negate with [\"NOT\", cond]. Valid operators depend on the field's data type (e.g. exact, in, isnull, exists, icontains, gt, gte, lt, lte, range); pick an invalid one and Paperless names the valid ones. Examples: '[\"Amount\", \"gte\", 100]', '[\"AND\", [[\"Due\", \"lt\", \"2026-10-01\"], [\"Paid\", \"exact\", false]]]'. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_filename / description
      Previous value: -"Filename pattern (whole name must match), * wildcards allowed"New value: +"Filename pattern (the whole name must match, case-insensitive), * wildcards allowed. Works for every trigger type: at consumption start it checks the incoming file's name, for types 2–4 the document's original file name."
    • changedInput schema / properties / triggers / items / properties / filter_has_all_tags / description
      Previous value: -"Document has ALL of these tags"New value: +"Document has ALL of these tags. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_any_correspondents / description
      Previous value: -"Correspondent is one of these"New value: +"Correspondent is one of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_any_document_types / description
      Previous value: -"Document type is one of these"New value: +"Document type is one of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_any_storage_paths / description
      Previous value: -"Storage path is one of these"New value: +"Storage path is one of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_not_correspondents / description
      Previous value: -"Correspondent is none of these"New value: +"Correspondent is none of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_not_document_types / description
      Previous value: -"Document type is none of these"New value: +"Document type is none of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_not_storage_paths / description
      Previous value: -"Storage path is none of these"New value: +"Storage path is none of these. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_not_tags / description
      Previous value: -"Document has NONE of these tags"New value: +"Document has NONE of these tags. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_has_tags / description
      Previous value: -"Document has ANY of these tags"New value: +"Document has ANY of these tags. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / filter_mailrule / description
      Previous value: -"Only documents fetched by this mail rule ID"New value: +"Consumption-started (type 1) triggers only: only documents fetched by this mail rule ID"
    • changedInput schema / properties / triggers / items / properties / filter_path / description
      Previous value: -"Path pattern, * wildcards allowed"New value: +"Consumption-started (type 1) triggers only: consume-folder path pattern, * wildcards allowed"
    • changedInput schema / properties / triggers / items / properties / matching_algorithm / description
      Previous value: -"Content match for `match`: 0=none, 1=any word, 2=all words, 3=exact, 4=regular expression, 5=fuzzy word"New value: +"Content match for `match`: 0=none, 1=any word, 2=all words, 3=exact, 4=regular expression, 5=fuzzy word. Checked for types 2–4 (added, updated, scheduled) only — not at consumption start, when the document has no content or metadata yet."
    • changedInput schema / properties / triggers / items / properties / sources / description
      Previous value: -"Consumption sources: 1=consume folder, 2=API upload, 3=mail fetch, 4=web UI (default [1,2,3])"New value: +"Consumption-started (type 1) triggers only: consumption sources 1=consume folder, 2=API upload, 3=mail fetch, 4=web UI (default [1,2,3])"
  2. First observedv3.2.0

TDQS

A4.1/5.0
Behavior4/5

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

Annotations only declare readOnlyHint=false and destructiveHint=false. The description goes further by disclosing execution semantics (actions run in order, ANY trigger fires the workflow) and the server-side validation rules that will cause a rejection, which an agent cannot infer from the schema alone. It does not describe what is returned or whether the created workflow is immediately enabled beyond the schema's default note.

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?

Front-loaded with purpose, then execution semantics, then a compact example, then the two rejection rules. Roughly four dense sentences with no filler. The example is worth its space, though the final constraint sentence partially duplicates schema text.

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

Completeness4/5

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

For a create tool with no output schema, the description supplies the payload construction model, cross-object validation rules, and run semantics — enough to invoke correctly. It omits what the call returns (e.g. the new workflow's ID) and any note on default `enabled` state, which are minor gaps given annotations cover the safety profile.

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

Parameters3/5

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

Schema description coverage is 80%, so the schema already carries most parameter meaning. The worked example (password-strip workflow: triggers [{type: 2, sources: [2]}] + actions [{type: 5, passwords: ["secret"]}]) adds concrete shape guidance, but the action-7/action-8 constraints it restates are already written in the schema's `type` descriptions, so it adds little net-new parameter meaning.

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?

States a specific verb+resource ("Create a complete workflow") and immediately clarifies the composite nature: triggers AND actions in one call, with the run-order semantics ("runs its actions in order whenever ANY of its triggers matches"). This clearly distinguishes it from the sibling create_workflow_action / create_workflow_trigger tools that build the same pieces one at a time.

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

Usage Guidelines4/5

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

The phrase "in one call" and the pointer that triggers/actions use "same fields as create_workflow_trigger / create_workflow_action" implicitly routes the agent to this tool for atomic workflow creation. It also states hard preconditions (remote OCR action requires a consumption-started trigger; apply-AI-suggestions requires a non-consumption-started trigger). It stops short of an explicit when-to-use-this-vs-the-siblings statement, so it is clear context rather than full routing guidance.

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

Deploy Server

Other Tools