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

Goes well beyond the readOnly/idempotent annotations: it discloses that execution happens on a hosted Cloudflare Worker so inputs leave the client, that the Worker is designed not to store inputs/results/tokens, that no further network call occurs and the map isn't forwarded, plus latency, size caps (50 texts, 200k chars each, 10k map entries) and the INPUT_TOO_LARGE error. It also flags the sensitive nature of the output and warns not to send it back to the LLM.

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?

Strongly front-loaded with WHAT/WHEN/WHEN NOT/COST/RETURNS labels that make it scannable despite its length. It is dense and repeats a few facts already in the schema (size limits, example map), so it is a touch longer than strictly necessary, but no section is filler.

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 nested-object, security-sensitive tool it covers everything an agent needs: source of the map, reversibility limits, replacement semantics, pagination, error case, output shape (restored_document rows with output_text, replacements, unknown_placeholders) and a worked example — even though an output schema exists.

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 already 100%, yet the description adds real semantics: placeholderMap should be passed as meta.placeholder_map unchanged, keys are matched exactly/case-sensitively with longest-key-first in a single pass so restored values are never re-replaced, and it explains what unknown_placeholders means and why an LLM-rewritten placeholder won't restore. It also clarifies that limit/offset page whole documents (default 10), which the schema only partially conveys.

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?

Opens with a specific verb+resource: putting original values back into texts containing placeholders like [EMAIL_1]. It explicitly names the source of the map (pii_redact_text / pii_guard_documents) and its inverse siblings (pii_redact_text, pii_scan_text), so an agent can distinguish it from every other PII tool without opening a schema.

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?

Has explicit WHEN (after an LLM answered a sanitized prompt, before showing the user) and WHEN NOT (not for redacting or scanning; masked/hashed values are not reversible — only placeholder style is). It names the concrete preconditions (redactionStyle "placeholder", returnPlaceholderMap=true), leaving nothing to inference.

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