keil_get_last_error

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

No annotations are provided, so the description carries the full burden. It usefully discloses that target hardware is not accessed, which suggests a side-effect-free read of cached state. It does not say whether reading consumes or clears the error, how stale the record may be, or what happens when no error has been recorded, leaving meaningful behavioral gaps.

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?

A single sentence that front-loads the action and includes only one qualifying detail. Nothing is padded and nothing obvious is buried.

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?

A dedicated output schema means return values need no explanation here, and a no-argument getter is inherently simple. The only remaining gap is lifecycle behavior — whether the error is consumed on read and its retention window — which keeps it short of full completeness.

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 tool takes zero parameters, and the description correctly requires no argument details. With no inputs to disambiguate, the baseline of 4 applies.

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 has a specific verb and resource: 'Return the latest recorded operation error'. This distinguishes the tool well from siblings that read state, registers, or build logs. It never explicitly differentiates itself from the error-related siblings such as keil_parse_build_errors or keil_get_session_log, so the distinction is implied rather than stated.

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 phrase 'without touching target hardware' implicitly tells the agent this error can be retrieved safely when the target is unavailable, halted, or not connected — useful context. However, there is no explicit statement of when to prefer this tool over session logs or build error parsing, and no mention of prerequisites.

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