Open Natural Gas: Georgia natural gas rates

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

Annotations already declare readOnly/idempotent/non-destructive, but the description adds real value beyond them: the cost formula, the fact that gated rows are returned unranked with a reason, that advertised deals are returned separately with evidence, and that it does not enroll anyone. It stops short of describing ordering guarantees or how the annual estimate is derived beyond the formula.

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?

Front-loaded with when-to-use triggers and then behavior, and nearly every clause carries information (scope, formula, unranked rows, deal caveat). The opening is one long run-on enumeration of intents, which is dense but not 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?

No output schema exists, so the description must carry return semantics — and it does: ranked rows by estimated annual cost, unranked gated rows with reasons, and a separate set of advertised deals with promo codes, conditions and evidence. An agent knows both what it gets and how to present it.

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%, so both parameters are already documented in the schema. The description reinforces usage ('at 900 therms', 'plans of one type') and explains how therms feeds the cost formula, but adds no format or edge-case detail beyond the schema, so the baseline of 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?

States a specific verb (compare/rank) and resource (Georgia natural gas marketer plans) with an explicit scope: residential Atlanta Gas Light customers, ranked by estimated annual cost from PSC filings. The exhaustive 'do not use for' list (electricity, commercial, Atmos, municipal, Regulated Provider, other states, wholesale) makes it unmistakable against siblings like marketer_profile and rate_history.

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?

Front-loads concrete user intents that select this tool ('cheapest natural gas in Georgia', 'what would I pay at 900 therms', promo codes) and gives explicit exclusions with the domains it must not be used for. It also instructs how to present results relative to advertised deals rather than declaring a PSC winner outright.

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

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

Annotations already declare readOnly/idempotent/non-destructive, and the description goes well beyond them: it specifies the return contents (filed plans, verified deals with promo codes and evidence, 24-month complaint counts, Google rating), states counts are 'as published; not scored or combined,' and imposes an output policy (no verdict, grade, or recommendation; rank/price are not quality measures). That is substantive behavioral context an agent cannot get from the annotations.

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?

Front-loaded with the trigger and the return payload before the policy and exclusion clauses, and the example queries earn their place by anchoring real user phrasing. It is dense and slightly overlong (the parenthetical examples and the anti-verdict policy could be tightened), but no sentence 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?

With no output schema, the description carries the return-value burden and does so fully: it enumerates the plan/deal/complaint/rating payload, notes the 24-month complaint window, and flags the evidence attached to deals. An agent has everything needed to call it and to interpret the response.

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% and the single parameter already documents partial matching and the ambiguous/unknown-name fallback, so the schema carries the load. The description only reinforces 'by name' and the single-marketer scope; it adds no matching syntax or format detail beyond the schema. 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?

Names a specific resource (one Georgia natural gas marketer) and enumerates exactly what it covers: reviews/Google rating, PSC complaints, plans and rates, contract terms, promo codes. It explicitly separates itself from compare_rates for multi-marketer comparison, so an agent can route correctly without opening the 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?

Gives an explicit trigger ('when someone asks about one Georgia natural gas marketer by name'), concrete example queries, and a hard exclusion list: multi-marketer ranking (use compare_rates), electricity providers, Atlanta Gas Light, Atmos Energy, municipal systems, and non-Georgia companies. When-to-use, when-not-to-use, and the alternative are all stated.

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

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

Annotations already declare a safe read-only, idempotent operation, so the description is free to add the traits that actually matter: the data source (PSC filings), the temporal boundary (since January 2019), the return shape (month-by-month cheapest and median estimated annual cost), and the caveat that gate-withheld months are listed rather than silently skipped. That data-quality disclosure is exactly the kind of context an agent cannot get from annotations.

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?

Two sentences, both front-loaded: the usage triggers lead, the return/data-caveat clause follows, and the exclusions close. Despite its length every clause carries actionable information (sibling routing, data provenance, withholding behavior, out-of-scope queries) with no 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?

With no output schema, the description still tells the agent what comes back (month-by-month cheapest and median estimated annual cost, across a stated date range) and warns about gaps in the series. Combined with the explicit scope boundaries, an agent has everything needed to call it correctly or to route elsewhere.

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%, so the baseline is 3 and the schema already documents formats and the four-value enum. The description adds marginal meaning on top, clarifying that results are for a single plan type at a time and giving the semantic gloss "fixed = standard 12-month", plus noting the marketer filter is optional.

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 concrete verb and resource ("how Georgia natural gas rates have changed over time") and enumerates the user questions it answers, including per-marketer price movement. It explicitly contrasts with compare_rates for current prices, so an agent can separate it from its sibling without opening the 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?

Provides both a positive routing clause ("Use this when someone asks...") with four concrete triggers and an explicit negative list: current-month pricing/ranking -> compare_rates, forecasts, electricity, Atmos/city systems, other states, wholesale/Henry Hub. This is the strongest form of when/when-not guidance.

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