Skip to main content
Glama

reparent_blueprint_safe

Destructive

Reparent a Blueprint with guarded rollback: preflight snapshot, optional variable collision renames, post-check verification, and auto-rollback on failure. Run dry_run=true first.

Instructions

Guarded, rollback-capable Blueprint reparent -- PREFER THIS OVER the raw reparent_blueprint, which has none of these guards. It adds: a pre-flight snapshot (parent chain, own-variable count, CDO component count); an OPTIONAL variable de-collision rename (rename_colliding_vars) for own variables whose name collides with a member the NEW parent already declares, which a plain reparent would otherwise clobber; the delta-serialization compile flags, without which a Blueprint with more than ~10 type-pinned variables CRASHES the editor during reinstancing; the same SCS-root reconciliation the Blueprint editor itself does, so a silently renamed component ("Mesh" -> "Mesh1") is reported instead of hidden; an invariant post-check (parent matches, no compile error, own-variable count did not shrink); and, on failure, an in-memory auto_rollback that restores the original parent, restores the component names and recompiles. It never saves to disk. This is a genuinely destructive multi-step transaction: run it with dry_run=true first on any real fork -- that reports the plan and the would-rename set without mutating anything.

Returns: { blueprint_name, blueprint_path, old_parent, new_parent, dry_run, reparented, renamed_variables:[{from,to}], pre_variable_count, post_variable_count, pre_component_count, post_component_count, pre_scs_node_count, post_scs_node_count, scene_root_reconciled, scene_root_reconciled_post_compile, default_scene_root_removed, sparse_class_data_conform_prepared, compile_messages:[{severity,text}], compile_error_count, compile_warning_count, renamed_components:[{guid,from,to,matched_by}], removed_components:[{guid,name}], components_without_guid, colliding_components:[{name,own_class,parent_class}], child_only_components:[{name,class}], compile_status, parent_match, compile_ok, variables_intact, components_intact, post_check_ok, rolled_back, rollback_complete, rollback_restored_components:[{guid,name}] (snapshotted nodes whose name MATCHES again after the rollback), rollback_unrestored_components:[{guid,expected_name,actual_name}], rollback_compile_messages:[{severity,text}], regressions:[string], saved_to_disk (always false -- this tool never saves), persistence_note, verification, verification_checks, success }

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
forceNoOptional. This command is REFUSED while a Play-in-Editor session is running: it either compiles a Blueprint (the reinstancer can retain a world, and CheckForWorldGCLeaks is Fatal -- it KILLS the editor) or writes to disk (Editor.AllowSavingAssetsDuringPIE=0 makes the engine refuse the write, so it would report success without writing a byte). Stop PIE and retry instead; force=true runs it anyway, at your own risk.
dry_runNoOptional (default false): analyze + report the plan (snapshot, would-rename vars, component collisions) without mutating.
auto_rollbackNoOptional (default true): on post-check failure, restore the original parent, restore the SCS node names and recompile.
blueprint_nameNoBlueprint asset name or path to reparent.
blueprint_pathNoAlias of `blueprint_name` (same value, same meaning). Accepted because sibling tools spell this parameter differently; prefer `blueprint_name`.
new_parent_classYesNew parent class name or path (native /Script, /Game/.../BP.BP_C, or a BP asset path -> its GeneratedClass).
reconcile_scene_rootNoOptional (default true): run the editor's own post-reparent conform pass -- ValidateSceneRootNodes + PrepareToConformSparseClassData + a second pass after the compile. Setting it FALSE disables all three; it exists only as a bisect escape hatch.
rename_colliding_varsNoOptional (default false): before reparenting, rename own NewVariables that collide by name with a member on the new parent.
treat_component_renames_as_regressionNoOptional (default false): count a silently renamed SCS component as a post-check regression (which triggers auto_rollback).

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv1.1.0

TDQS

A4.5/5.0
Behavior5/5

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

Annotations already flag destructiveHint=true and idempotentHint=false, but the description goes far beyond them: it enumerates the guards (snapshot, de-collision rename, delta-serialization flags that prevent an editor crash, SCS reconciliation, post-check invariants), the auto_rollback failure semantics, and critically states 'It never saves to disk' plus that it is an in-memory multi-step transaction. This is exactly the context an agent needs before firing a destructive op.

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

Conciseness2/5

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

The prose body is one enormous run-on sentence with deep semicolon nesting, and it is followed by a ~40-field return dump. Front-loaded value proposition is good, but the return-field enumeration ('scene_root_reconciled_post_compile, default_scene_root_removed, sparse_class_data_conform_prepared...') is far too long for a description and belongs in an output schema (which is absent). Signal-to-noise is poor on the back half.

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?

There is no output schema, so the description must (and does) explain the return shape and semantics — including rollback fields, compile counts, and saved_to_disk. It covers the failure/rollback path, the dry-run path, and the PIE precondition. The only shortfall is that enumerating every return field bloats the description rather than summarizing, and there is no mention of rate limits or concurrency, but coverage is otherwise strong.

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 description coverage is 100% and schema descriptions are already rich (force explains the PIE refusal, dry_run default, auto_rollback semantics, rename_colliding_vars, reconcile_scene_root). The description adds a coherent narrative correlating the flags to the guards, but does not add per-parameter detail beyond the schema. Baseline for 100% coverage is 3; the integrating narrative raises it slightly.

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 ('Blueprint reparent') plus a distinct value proposition (guarded, rollback-capable) and explicitly names the sibling it replaces ('PREFER THIS OVER the raw reparent_blueprint'). An agent can distinguish reparent_blueprint_safe from reparent_blueprint and verify_reparent_success 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.

Usage Guidelines5/5

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

Gives an explicit preference rule ('prefer this over the raw reparent_blueprint'), a mandatory workflow ('run it with dry_run=true first on any real fork'), and a hard precondition ('REFUSED while a Play-in-Editor session is running' in the force param). When-to-use and when-not-to-use are both covered.

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