Scan Datasource

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

Annotations already declare idempotentHint=true, destructiveHint=false and readOnlyHint=false, but the description adds substantive context beyond them: what profiling actually inspects (samples, uniqueness, date formats, small value sets), how keys are inferred (by name, then confirmed by value containment), and crucially what survives a re-run (existing Terms and reviewed relationships). The one remaining gap is cost/scale — no warning that profiling samples rows and may be slow on large schemas, which matters since readOnlyHint=false signals it does persist inferred keys.

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?

Two sentences, front-loaded with the core action, with no filler. The semicolon-chained clauses pack a lot of information but remain parseable, so the density costs only a little readability.

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 scanning/mutation tool whose annotations cover safety and whose output schema covers return values, the description supplies what matters: scope of the scan, inference methodology, and preservation guarantees. It stops short of stating runtime cost or required datasource permissions, which is the only notable omission.

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 100%, so the schema already documents all three parameters and the baseline of 3 applies. The description does add modest meaning by spelling out what the `profile` flag causes (sampling and value-level tests) and by implying `schemas` scopes the read, but it offers no guidance on the cost trade-off of disabling profiling or on when `datasource` can be omitted.

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 names a specific verb set (read/re-read, profile, infer, confirm) and a concrete resource (tables, columns, declared keys, column values). It is unmistakably about schema scanning and key inference, but it never contrasts itself with siblings such as check_readiness, add_datasource, or get_modeling_worklist, so an agent must still infer which of the 30+ tools is the right entry point.

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?

"Read (or re-read)" implies the two triggering situations — first onboarding of a datasource and a later refresh — and the closing sentence signals re-running is safe for curated content. However, no alternative tool is named and no exclusion is stated (e.g., when to prefer check_readiness or add_datasource first), leaving routing to inference.

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