QC-002 Grover presented-oracle verifier. oracle is a 0/1 string of length 2..8; marked is the unique 1-index. Does not search the space. Floats refused.
| Name | Required | Description | Default |
|---|---|---|---|
| marked | No | ||
| oracle | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full behavioral burden. It usefully discloses that the tool does not search and refuses floats, which are meaningful behavioral constraints. However, it does not mention side effects, return value/format, or error behavior for invalid inputs such as malformed or missing oracle strings.
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?
Three short sentences deliver the tool's role, parameter constraints, and behavioral caveats with no filler. The most important categorization ('verifier') is front-loaded, followed by concise parameter semantics and a clear non-search disclaimer.
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 simple two-parameter tool, the definition covers the inputs and one key behavioral detail. Still, with no output schema and no annotations, the description should state what the verifier returns (e.g., boolean valid/invalid) and how it handles invalid or missing arguments. The agent can infer some of this from the name and constraints, but it is not explicit.
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?
The input schema provides bare string types with no descriptions, so the description is the only semantic source. It adds crucial meaning: oracle must be a 0/1 string of length 2..8, and marked is the unique 1-index. It would be stronger with an explicit format for marked (e.g., integer within 1..8) and requiredness, but it substantially compensates for the 0% schema coverage.
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 clearly identifies this as a Grover 'presented-oracle verifier' and explains the input roles: oracle is a 0/1 string and marked is the unique 1-index. It also distinguishes the tool from a search operation by explicitly saying it does not search the space. However, it stops short of stating the precise verification criterion (e.g., whether it asserts that the oracle has exactly one marked position and that index matches marked).
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 statement 'Does not search the space' gives an explicit exclusion: this is not a tool for executing Grover search. That is useful routing information. But no positive condition is given for when to select this verifier over the many sibling verify_* tools, and no alternative tool is named.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Many verify_* tools are distinct, but there are overlapping clusters: math_court duplicates execute_2local_hamiltonian, route_spin_glass_manifold, and the quantum verifiers; project_affine_key, expose, verify_presented_pair, and the optional affine exposes in other tools blur together; weather and geometry tools also overlap. The detailed descriptions help a human, but an agent would likely struggle to choose between equivalent-seeming entry points.
Naming is mostly snake_case but otherwise inconsistent: some tools use dotted prefixes (atc.*, twin.robotics.*, weather.*), some use bare verbs (expose, lattice_op, math_court), some use noun phrases (corpus_bonds, feeds_catalog), and others mix prefixes with verbs (ide_rebuild_mesh, umc_resume). The verify_* family is consistent, but the overall set has no single predictable verb_noun pattern.
49 tools is far above the typical well-scoped server size and includes multiple near-duplicate paths to the same law (math_court, execute_*, route_*, verify_*). While not quite 50+, the count still feels like a sprawling kitchen-sink rather than a deliberate minimal surface.
The toolset covers a surprisingly wide range: QC verifiers, QMA laws, affine projections, corpus reads, weather, UMC state, and robotics IK. However, there are notable gaps for such a broad surface: no general court case lifecycle beyond expose/seal, no corpus content search, and no way to manage or update sealed artifacts; several areas have only entry-point coverage.