Sourced, dated answers for agents (FR/EU rules, e-invoicing, Qualiopi) + market math on your data
Scored across 3 tools
Each tool targets a clearly distinct function: 'ask' handles external factual Q&A with sources, 'calculate' performs deterministic computation on user-provided candles, and 'feedback' rates a previous 'ask' response. There is no meaningful overlap in purpose.
All names are single-word, lowercase English tokens with no underscores or camelCase, giving a consistent visual style. The only minor deviation is that 'ask' and 'calculate' are verbs while 'feedback' is a noun.
Three tools are well-scoped for this API: one to query verified facts, one to run calculations on user data, and one to close the feedback loop. Each earns its place without redundancy or obvious missing essential tool.
The ask/feedback lifecycle and deterministic calculation are covered. Minor gaps exist: feedback cannot be given on 'calculate' results, and there is no direct way to retrieve a previous answer or request status by request_id, though agents can work around this.
Pose une question factuelle et reçoit une réponse vérifiée avec ses sources, sa confiance, sa date d'expiration et un request_id. Domaines facultatifs : actualite, droit, entreprise, facturation, finances, formation, general, impots, logiciel, prix, reglementation, vie_quotidienne. Facultatif : context, la tâche que vous êtes en train de faire (sans données personnelles).
| Name | Required | Description | Default |
|---|---|---|---|
| domain | No | ||
| context | No | ||
| question | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and openWorldHint=true, so the external-lookup, non-mutating nature is partly covered. The description adds that answers are verified and carry confidence and expiration, plus a no-personal-data caution for context, but does not elaborate on latency, source freshness, or failure behavior.
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?
Purpose is front-loaded in the first sentence, followed by the domain value list and the context note. The long domain enumeration is functional rather than filler, though the whole is slightly dense.
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?
An output schema exists, so return values need not be re-explained, yet the description helpfully previews them. With domain values and context documented, an agent has enough to call the tool correctly despite the 0% schema coverage.
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 0%, so the description must compensate, and it does: it enumerates the accepted domain values and explains what context is for (the current task, without personal data). Only 'question' is left to common sense, which is reasonable.
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 (pose une question) and resource (question factuelle), and even summarizes the returned artifact (réponse vérifiée avec sources, confiance, expiration, request_id). The purpose is clear, but there is no explicit differentiation from the sibling tools calculate and feedback.
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 word 'factuelle' implicitly scopes usage to factual questions needing verified answers, which is useful direction. However, there is no statement of when to prefer this over calculate or feedback, and no exclusions or prerequisites are given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Calcul déterministe sur VOS bougies (304 ne fournit aucune donnée de marché). calculation : ichimoku-rvol (Ichimoku confirmé par le volume relatif), regime (tendance ADX, volatilité ATR, direction) ou position-size (taille de position pour un risque fixé). params : le corps JSON de POST /v1/calc/, par exemple {"candles": [{"time", "open", "high", "low", "close", "volume"}, …]} ; voir /llms.txt. Résultat : méthode publiée, valeurs, horodatage. Ce n'est pas un conseil en investissement.
| Name | Required | Description | Default |
|---|---|---|---|
| params | Yes | ||
| calculation | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already mark this readOnly and idempotent, so the safety bar is lower. The description still adds meaningful context beyond them: the calculation is deterministic, 304 supplies no market data (you must pass your own candles), the result contains the published method/values/timestamp, and it is not investment advice.
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 core purpose, then mode definitions, then the params format, then the return/disclaimer note. Dense but every sentence carries information; no redundant restatement of the schema.
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?
An output schema exists so return values need not be described in depth, and the description still sketches the result. It covers the deterministic/privacy framing, the enum, and a params example, leaving only per-calculation parameter detail as a gap.
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 0%, so the description carries the full burden. It maps the 'calculation' enum to concrete meanings and shows a 'params' example with the candle object shape (time/open/high/low/close/volume), plus a pointer to /llms.txt. But it only illustrates one of the three calculations' bodies and does not specify the params each mode requires, leaving gaps for a schema with open nested objects.
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 and resource ('Calcul déterministe sur VOS bougies') and enumerates the exact three calculation modes (ichimoku-rvol, regime, position-size), each glossed. An agent can distinguish this computational tool from the sibling 'ask'/'feedback' tools immediately.
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 explains what each calculation mode produces (Ichimoku confirmed by relative volume; ADX trend/ATR volatility/direction; position size for a fixed risk), which gives clear context for choosing a mode. However it never states when to prefer this tool over the 'ask' sibling, so no explicit alternatives or exclusions are given.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Dit si une réponse obtenue avec « ask » vous a servi. request_id : celui de la réponse. useful : true ou false. issue (facultatif) : wrong (fausse), outdated (périmée), incomplete, bad_source (source insuffisante), off_topic (hors sujet), contradiction (une autre source dit autre chose) ou other. comment (facultatif) : ce qui manquait ou ce qui était faux.
| Name | Required | Description | Default |
|---|---|---|---|
| issue | No | ||
| useful | Yes | ||
| comment | No | ||
| request_id | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations declare readOnlyHint=false and idempotentHint=true, so mutation and repeat-safety are already covered. The description adds useful semantics for the issue field and notes optionality of two params, but doesn't state permission requirements or the effect of sending feedback on downstream data. This is adequate additional context but not rich.
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 and front-loaded, opening with the core purpose before parameter notes. The long enumeration of issue values is necessary for semantics but contributes to density. No filler sentences.
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 an output schema exists, return values need not be described. With zero schema description coverage, the description supplies the missing parameter semantics and links to the ask tool. It is nearly complete, lacking only any note on permissions or consequences of submitting feedback.
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 0%, so the description must compensate. It explains request_id refers to a prior « ask » response id, useful is true/false, and enumerates all seven issue enum values with glosses, plus optionality of issue and comment. This meaningfully adds to the bare schema, though comment could be slightly more specific.
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 states a specific action tied to a specific upstream tool: telling whether a response obtained with « ask » was useful. It clearly connects to its sibling ask and is distinguishable from calculate. It does not explicitly name a verb like 'submit feedback', but the intent is unambiguous.
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 clarifies this operates on responses from a specific sibling tool (« ask ») and requires the request_id from that response. It does not discuss timing or when not to send feedback, but the contextual link to ask is strong guidance for invocation.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
askcalculatefeedbackMarket knowledge layer for AI agents: competitors, opinions and regulations shaping your market.
Connect AI agents to licensed financial intermediaries in France: insurance, credit, wealth.
French legal helpers for agents: e-invoice, L441-10, SIRET/IBAN. $0.01 USDC x402
Evidence-backed capital-change intelligence and sourced financial data for AI agents
Add one secure layer between your agents and this server.
