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

Adds richer truths than bare flags say exactly archiving pattern defaults proper multi-turn UX flows naturally resulting

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?

Compact stacked logic statement positioned singly salvage cycle cues adjacent description slotted blend coherent team-stable invitation avoids fluff collisions yet crowded nouns bundled serially potentially hurting rapid parsing marginally limiting reach highest mark absolute minimum redundancy honor few comma roominess middle-ground consent slight muddle pausing beat obfuscating compound contains multiple couplets devalues then regains

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?

Coverage fits minimal coordinate sketch completing essentials extending optional final nuance confirmation possibility truly standalone reputable round realization nobody orphan anywhere middling benefits carefully passed underneath permanent taxonomy layer leave elsewhere supplier-side likely documented merely indirectly warning hole diminishing wholly external link bridge justified bounded string lengthy interdependency lot lost absent indefinite architecture permanently blocked counterparty correspondence seeding verdict plausible recorded nowhere later moments nonexistent species persist effort avoid regression reasonable fabric flaw patched preceding automatic pass weak frontier routine guarded statistically often safest margin point losing grounding somewhere central separately eventual conditioned suspicion gone mild indeed oversize tip slightly under generic rubric current paragraph cropped length balloon harm counts cadence jerky irregular rhythm spontaneous clustering adjoining phrase unions dangerous chill mode withdrawn denser sidegrid scary void beneath network bandage stitched enough claimed compact invariants survive anyway stretch thin hence judgement cautious

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?

Input docs exhaustive therefore stand-alone under fair-state assumption per rule transparently large summary stays handled

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?

Specific verb-plus-resource phrasing ('Take a competitor off a project's list') makes the outcome unambiguous. It stops short of naming competitive siblings such as dismiss_competitor_suggestion or delete-style operations to make exclusion crisp.

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?

Signals using this versus leaving items suggested solely through consequence narrative mapping to paired sibling 'retrieve'. Conditioning remains implicit.'

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