scene_set_class

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

With destructiveHint=true declared, the description adds genuinely mitigating context: it writes a NEW .gil into the workspace and leaves the official save untouched, so the agent can reason about risk. It also discloses that health/attack/defence/stamina are base values multiplied by a growth curve at the given level, and cites play verification. It never says what happens to fields that are omitted (all 13 params are optional), which is the main behavioral blind spot.

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?

Purpose and key fields are front-loaded in the first sentence, which is good structure. But the body is bloated with dated playtest forensics (specific meter distances, a 12 m tower, a 2.4 m wall, verification dates) that convey semantics at far more length than an agent needs, pushing the definition past 250 words.

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 13-parameter mutating tool with no annotations beyond hints and no output schema, the description supplies rich domain context: class-based ability model, combat-attribute math, camera sourcing, and output safety. The gaps are the unexplained mapPath and outputName, and the absence of any statement about defaults or required arguments when errors occur.

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 coverage is 0%, so the description carries the full burden and does most of it: allowJump/allowSprint/allowClimb/allowGlide are explained with concrete observed effects and units (17.4 m vs 10.5 m, 2 m/s glide), cameraTemplateId is tied to scene_list_cameras, name is tied to disambiguating among several classes, and level/health/attack/defence/stamina are explained relationally. mapPath and outputName are never mentioned, leaving two of thirteen parameters unexplained.

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 first sentence names a specific verb (change) and resource (a class's abilities/camera) and enumerates the exact fields it controls, so an agent knows precisely what the tool mutates. It also draws the boundary against graph-based work ('abilities and camera belong to the class, not to node graphs'), which separates it from the graph-building siblings.

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 real situational guidance: use it when a side-scroller or top-down game needs the right camera and unused abilities switched off, and it warns that a new map's default 10000 health makes the player invincible so games with losing fights must tune health. It names scene_list_cameras as the source for cameraTemplateId, but offers no explicit exclusions or pointer to scene_list_classes / scene_set_level_settings as alternatives.

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