rupa_rantai

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

Annotations already communicate readOnlyHint=false and destructiveHint=true; the description adds valuable behavior beyond that: the chain halts at the first failure ('RANTAI BERHENTI DI KEGAGALAN PERTAMA'), per-op timing is intentionally not measured, and only whole-chain time is reported. It also makes the 'identical results' claim explicit and keeps performance claims empirical. This is genuinely supplementary, not contradictory.

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?

The description is longer than a one-liner, but it earns its length by establishing why the tool matters (benchmark figures), how execution differs (fail-fast, no per-op timing), and when not to use it. The main purpose is front-loaded in the first sentence. The benchmark details are slightly over-elaborate but acceptable; this is near the upper bound of 'no wasted sentences.'

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?

Given that there is no output schema, an agent is left uncertain about the return/reported result beyond the statement that whole-chain time is reported. The description covers batching eligibility, failure behavior, and consistency, which goes a long way, but it never states what happens when an op is not describing return values or how successful results are delivered. The absence of per-op timing and return shape leaves a gap for a tool of this complexity.

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?

Second-dimension schema coverage is 100%: both `ops` and `ruang` have complete descriptions, so the baseline is 3. The description content adds no direct parameter-level detail (e.g., no mention of the `ops` array format or `ruang` placement) but it also does not need to repeat schema information. It receives no credit beyond the schema.

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 description opens with an unambiguous verb and resource: 'Jalankan beberapa op Blender dalam satu proses.' It clearly defines the tool as a batching mechanism and contrasts it with repeated single-op calls ('Enam op satu per satu makan 22,2 detik'), which distinguishes it from the many individual `rupa_*` siblings even without naming them. A full 5 would be justified if it explicitly named the alternative tools such as `rupa_lod` or `rupa_ukur`.

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?

The description gives an explicit when-not condition: 'YANG TIDAK BISA DIRANTAI: urutan yang punya KEPUTUSAN di tengah' and tells the agent to split these into separate chains ('Pecah jadi beberapa rantai'). It also explains the failure-stop approach but stops one step short of naming which single-op tool to use as an alternative for a chain that can't be batched.

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