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

Annotations cover the safety profile (read-only, closed-world, non-destructive), and the description adds substantial behavior far beyond them: the seal check, pinned-input content validation with changed files refused by name, the exact selection rule for which hypotheses run, the guarantee that answered/settled/cancelled hypotheses never re-run, and the budget-slicing loop. This is unusually rich disclosure for a resumption tool.

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?

Three sentences, front-loaded with the core action and condition before descending into validation and scheduling detail. It is dense and clause-heavy, but essentially every clause carries a distinct behavioral fact, so little could be cut without losing information.

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 high-complexity, nested-object tool with five parameters, the description covers trigger, validation constraints, execution selection rules, continuation protocol and output equivalence. An output schema exists, so the brief return-value note is a bonus cross-reference rather than a necessity.

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 100%, so the schema already documents run, sequences, _files, _subject and budget_seconds in detail. The description adds only the slicing-loop semantics for budget_seconds ('call again with the returned file until nothing is pending'), which is useful but marginal against a fully documented schema — the baseline 3 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 first clause gives a specific verb and resource — 'Continue a research run from its file' — plus the triggering condition (interruption or failed test). It is clearly distinguishable from siblings like cancel_research_run, research_run_status and run_discovery_programme, the last of which it explicitly anchors itself to by output equivalence.

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?

It states when to use it: after an interruption or a failed test, and repeatedly with the returned file when budget_seconds is set. The resume-only framing implicitly excludes starting a fresh run, but it never names run_discovery_programme as the alternative for that case, so the routing guidance is clear but not exhaustive.

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