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

Annotations already declare read-only, idempotent, non-destructive behavior. The description goes well beyond this by disclosing mode-specific behavior (single-market vs basket), how size_usd is interpreted differently per mode (max spend vs target proceeds vs settlement notional), the internal 'walks the ladder' logic, and the warning about partial fills converting arbs into directional positions. This is rich behavioral context that annotations do not provide.

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 description is long (around 300 words) but every sentence carries necessary information. It is well-structured: purpose, requirements, single-market mode details, basket mode details, and a concluding usage recommendation. It is front-loaded with the core purpose and mandatory inputs. Slightly verbose in listing all return fields, but that detail is useful given there is no output schema. Overall, it earns a 4—not a 5 because it could be tightened without losing value, but it is far from wasteful.

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?

Given the tool's complexity (two modes, four params, no output schema), the description is remarkably complete. It lists the key return fields for both modes (top_of_book, vwap_fill_price, slippage_pp, verdict, thin_legs[], forced_directional_risk, etc.), explains the risk scenario (partial fills leading to unhedged positions), and provides usage thresholds. It leaves little ambiguity about what the tool does, what inputs it accepts, and what outputs to expect.

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 100%, but the description adds significant semantic value: it explains that `market` and `event` are mutually exclusive modes, clarifies the default side behavior in basket mode ('auto — sell if partition sum > 1, buy if < 1'), and gives a precise definition of `size_usd` for each mode ('max spend on buys, target proceeds on sells' vs 'settlement notional — shares per leg, each paying $1'). This goes far beyond the schema's one-line property descriptions.

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 opens with a specific verb+resource phrase: 'Realizable-vs-theoretical edge check against live CLOB order-book depth.' It clearly names the tool's function and distinguishes it from siblings by explicitly referencing `polymarket_arbitrage` and `polymarket_edges` as the tools it supports, and explaining it checks whether theoretical edges are actually capturable given order-book depth.

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 description gives explicit when-to-use guidance: 'USE THIS before acting on any polymarket_arbitrage SELL/BUY-EVERY-LEG signal or any polymarket_edges trade above ~$500.' It also explains why (theoretical overround on thin books is not capturable, partial basket fills convert arb into unhedged directional risk), effectively covering when-not-to-use by implying small trades below the threshold don't require it. It names the alternative tools it complements.

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