MPFB MCP

Official

Overview Schema Source Related Servers Discussions Admin

mpfb_get_target_stack

Read-onlyIdempotent

Read a MakeHuman character's applied shape keys and modeling targets without changing anything, including per-category modifiers, counts and left/right asymmetries.

Instructions

Read the modeling targets currently applied to a MakeHuman character - every shape key on its basemesh, and the modeling sliders they add up to. Changes nothing.

result has two views of the same facts, addressed differently.

targets - one entry per shape key: shapekey_name (the raw Blender name, and the address mpfb_set_targets' targets list takes), decoded_name (readable; a name over 60 characters is stored encoded, and only the raw one works as an address), value, kind (macro/expression/detail/unknown), and where the target sits in MPFB's index: section, category, side, polarity, listed_in_target_json, all null for a user target belonging to no category.
modifiers - the same information rolled up per target.json category, each {section, category, side, value} with value in -1.0..+1.0. This is the view MPFB's own sliders show and the one that feeds straight back into mpfb_set_targets' modifiers list. Only categories with a non-zero reading appear; never paged, never filtered by name_contains.

asymmetries lists the left/right categories whose two sides differ - which MPFB's own UI does not surface anywhere - and is what makes mpfb_symmetrize_targets an informed operation rather than a blind one. basemesh_name names the object the stack was read from.

include_macro (default false) adds the $md- phenotype shape keys, which are mpfb_get_macro_details' subject; include_expressions (default false) adds the !ex- expression ones. counts_by_kind reports both counts whether or not they were included.

The answer is capped. limit defaults to 200 and cannot exceed 1000. Read truncated and total_count before concluding a target is not on this character, and narrow with name_contains rather than paging blindly; counts_by_section says where the entries are.

This tool sees one thing MPFB itself cannot. MPFB excludes any shape key whose name contains "basis", not merely equals it, so a user target called basis-nose-widen is real, is deforming the mesh, and is invisible to every MPFB read and write. It is listed here with visible_to_targetservice: false, and it cannot be changed through mpfb_set_targets either.

A subject that cannot be resolved to a basemesh is an answer, found: false with a sentence, not an error. A character whose macro numbers and macro shape keys have drifted apart is reported as it is, not repaired.

Input Schema

TableJSON Schema

Name	Required	Description	Default
`name`	No
`limit`	No
`offset`	No
`include_macro`	No
`name_contains`	No
`include_expressions`	No

Output Schema

TableJSON Schema

Name	Required	Description	Default
No arguments

Schema Changelog

Changes observed during successful MCP inspections.

First observedv0.1.0Oct 3, 2026

TDQS

A4.6/5.0

Behavior5/5

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

Despite annotations already declaring readOnly/idempotent/non-destructive, the description discloses a great deal more: the 200-default/1000-max cap, the never-paged/never-filtered behavior of the modifiers view, the found:false-not-error contract, that drifted macro values are reported rather than repaired, and the invisible 'basis'-containing shape keys MPFB cannot see. This is exactly the kind of behavior beyond annotations the dimension rewards.

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 the core purpose and well-structured with bolded section headers. It is long and spends substantial prose enumerating result fields (targets, modifiers, asymmetries) that the existing output schema already defines, which is some duplication, though most sentences still carry semantic value.

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 six-parameter read tool with a rich output schema, the description covers the truncation contract, inclusion flags, filtering, edge cases (unresolvable subject, drifted macros, invisible keys), and the dual addressing model. An agent would not need to guess at call semantics or failure behavior.

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

Parameters4/5

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

Schema coverage is 0%, so the description must carry the burden, and it does well for include_macro, include_expressions, limit (default 200, cap 1000), and name_contains. However `name` and `offset` are never explained, leaving two parameters undocumented in both schema and description.

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 and resource: 'Read the modeling targets currently applied to a MakeHuman character - every shape key on its basemesh, and the modeling sliders they add up to.' It explicitly contrasts itself with siblings by naming the write counterpart (mpfb_set_targets), the macro-subject tool (mpfb_get_macro_details), and the downstream consumer (mpfb_symmetrize_targets).

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?

Gives clear operational context: read truncated/total_count before concluding a target is absent, and 'narrow with name_contains rather than paging blindly.' It also frames asymmetries as the input that makes mpfb_symmetrize_targets informed. It never states an explicit when-not-to-use or a strict alternative-selection rule, so it stops short of a 5.

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