propose_fix

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

With no annotations, the description carries the full burden and delivers the key traits: output is a diff/SQL proposal, 'never applied' (non-mutating), and deleted clusters are refused. It omits permission/rate-limit details, but for a proposal-only tool the core safety profile is disclosed.

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 opening sentence front-loads purpose and the non-applied nature; the dense subsequent text earns its place by mapping issue values to behaviors and parameter quirks. It's information-dense rather than padded, though the nested parenthetical structure is heavy.

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?

The output is a proposal (diff/SQL) and an output schema exists, so return values needn't be explained. The description covers the tricky parameters and safety behavior, leaving only a few auxiliary params (path, num_workers, spark_version) unaddressed.

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?

At 0% schema coverage, the description compensates well: it explains the issue enum, tells that delta_zorder overloads node_type_id as column list, that delta_vacuum uses minutes (default 168/min 7 days), resource_key as catalog.schema.table, and that bundle_yaml may be empty. It leaves path, num_workers, and spark_version undocumented, so not fully complete.

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?

States a specific verb and resource ('Propose a fix as a unified diff ... or reviewable SQL') plus a critical behavioral qualifier ('never applied'). This clearly distinguishes it from siblings like find_waste, cost_summary, and top_cost_drivers, which are analysis/reporting tools rather than fix generation.

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?

Enumerates issue types with their intent, tells the agent to pass cluster_id 'from find_waste' for cluster fixes, and notes a deleted cluster is refused. It establishes the routing from find_waste and the delta-issue path, but doesn't explicitly say when NOT to use it or name every alternative.

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