canvas_sync_courses

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

Despite annotations already declaring the safety profile, the description adds substantial context the agent could not infer: the operation is local-only and cannot be invoked from a remote cloud server, dry-run is the default, existing drafts are never overwritten, no schedule is created, and partial coverage/failed sources are possible outcomes. It even warns that 'failed sources do not mean no work,' which is exactly the kind of non-obvious behavioral caveat that prevents misreads of results.

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 content is dense and mostly earns its place, but the structure is poor: it opens with a semicolon fragment, then stacks constraints and caveats in hard-wrapped lines with no hierarchy. 'Remote cloud servers cannot write to your Mac' restates 'LOCAL STDIO ONLY' and could be merged, and the actual operation is never stated up front.

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?

An output schema exists, so return values needn't be described, and the safety/scope caveats are well covered. However, for an 8-parameter tool with zero schema descriptions, the absence of any explanation for course selection and the four limit parameters is a real completeness gap that the description should have closed.

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 description coverage is 0%, so the description carries the full burden, and it only covers two of eight parameters: output_root ('must be an absolute private coursework directory outside connector source') and apply (dry-run vs write). The six remaining parameters — course_ids, all_active, max_files, max_file_mb, max_total_mb, max_operations — get no explanation at all, leaving the agent to guess at scoping and throttling behavior from bare titles.

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 opening sentence gives a specific verb+resource ('Plan or update separate per-class folders') and immediately distinguishes the tool from write-capable siblings with 'never write to Canvas.' It is not maximally crisp because the purpose is buried in a fragment followed by a wall of constraints, and 'sync' never appears explicitly, but an agent can identify what the tool does.

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 apply semantics are spelled out clearly: 'Default apply=false discovers changes without local writes. apply=true preserves original bytes, versioned hashes, extraction previews, manifests...' — that is genuine when-to-use guidance for the primary mode switch. It falls short of a 5 because it never routes the agent to or away from any named sibling tool, and the 'LOCAL STDIO ONLY' precondition is stated but not connected to a fallback.

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