Public tenders (EU TED, Armenia) for AI agents: search, details, watchlists, alerts.
Scored across 4 tools
Each tool has a distinct role in the workflow: search_tenders finds tenders, get_tender retrieves details, add_to_watchlist creates monitoring, and get_deadline_alerts reads alert results. There is no meaningful overlap that would cause an agent to misselect.
All tools use snake_case and verb-led names, with get_ used consistently for retrieval tools. add_to_watchlist is slightly different due to the preposition, but it remains readable and predictable.
Four tools form a compact but coherent core surface for searching, inspecting, watching, and receiving alerts on tenders. The count is reasonable for the domain, though the set is slightly thin on watchlist management.
The core workflow is covered, but there are notable gaps around the watchlist lifecycle: no tool to list, update, or delete watchlists, and no way to mark alerts as read. The dashboard is offered as a workaround, which reduces but does not remove the missing API operations.
Save a query and filters as a watchlist for ongoing monitoring; new, amended and closing-soon matching tenders are then sent by email (Free plan: daily digest) and/or a signed HTTPS webhook (Pro plan). Returns watchlistId and how many open tenders match right now. Each call creates a new watchlist (the number per plan is limited); manage or delete watchlists in the dashboard.
| Name | Required | Description | Default |
|---|---|---|---|
| name | Yes | Watchlist name | |
| filters | Yes | Which tenders to watch: the same query and filters as search_tenders (q, countries, cpv, budget, deadlineAfter); filters that match every tender are rejected | |
| channels | No | Where alerts go: email (default) and/or an HTTPS webhook; webhooks need the Pro plan, Free gets a daily email digest |
| Name | Required | Description |
|---|---|---|
| hint | No | Note about the filters, e.g. which currency the budget range uses |
| watchlistId | Yes | |
| estimatedMatches | Yes | Open tenders matching the filters right now |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover non-readonly/non-idempotent behavior; the description adds substantial context beyond them: each call creates a new watchlist, plan-based count limits, Free-plan daily digest vs Pro webhook, signed HTTPS delivery, and rejection of filters that match everything.
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 core purpose is front-loaded in the opening clause, and the remaining sentences carry plan, delivery, and lifecycle constraints rather than filler. The dense semicolon-heavy style is slightly harder to scan but nothing is wasted.
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 and full annotation coverage, the description still supplies the operational context an agent needs: creation semantics, plan limits, alert delivery channels, and where to manage existing watchlists. Nothing material is missing.
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 schema already documents parameters thoroughly. The description still adds value by tying `filters` to the search_tenders semantics and explaining channel plan constraints (webhook requires Pro, email default on Free).
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 ('Save a query and filters as a watchlist') plus the downstream effect (ongoing monitoring, email/webhook alerts) and return value (watchlistId, current match count). This clearly distinguishes it from search_tenders, which is a one-off query tool.
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?
Frames the tool around 'ongoing monitoring' versus a one-time search, notes per-plan watchlist limits, and routes management/deletion to the dashboard. It stops short of an explicit 'use search_tenders instead when you just need results now' statement, but the context is clear.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List alerts from your watchlists for the last 30 days on tenders that are still open, nearest deadline first (up to 50). Does not mark alerts as read. Empty until you create a watchlist with add_to_watchlist.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| hint | No | |
| alerts | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Adds behavior beyond the annotations' readOnly/idempotent profile: it discloses the side-effect boundary ('Does not mark alerts as read'), the result cap (up to 50), the ordering, the 30-day and still-open filters, and the empty-state cause. These are exactly the traits an agent needs to interpret results without re-calling.
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 short sentences, fully front-loaded: what it returns and how it is scoped first, then the side-effect clarification, then the prerequisite. No filler or restatement of the title.
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 an output schema present, return values need no prose explanation, and the description still covers filtering, ordering, cap, idempotent non-marking behavior, and the prerequisite. Nothing an agent needs in order to call it correctly is absent.
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?
The tool takes zero parameters, which is the baseline-4 case; there is nothing for the description to disambiguate. The schema confirms an empty properties object, so no parameter semantics are missing.
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 (List) and resource (alerts from your watchlists) plus precise scope: last 30 days, tenders still open, nearest deadline first, capped at 50. This is clearly separable from get_tender and search_tenders, which query tenders directly rather than the user's watchlist alerts.
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 a concrete prerequisite that doubles as routing guidance: 'Empty until you create a watchlist with add_to_watchlist', pointing at the sibling needed to make this tool useful. It does not explicitly contrast with get_tender or search_tenders, so it stops short of full when/when-not coverage.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get the full card of one tender: status, deadline, budget, what is procured (subject), lots with prices and CPV codes, links to tender documents. Use the id from search_tenders results. On the Free plan, CPV codes, lots and document links are omitted.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | Tender id from search_tenders results |
| Name | Required | Description |
|---|---|---|
| id | Yes | |
| url | Yes | |
| hint | No | |
| lots | No | Omitted on the Free plan |
| buyer | Yes | |
| title | Yes | |
| budget | Yes | Estimated budget as a decimal string, e.g. "125000.00" |
| region | Yes | |
| source | Yes | |
| status | Yes | |
| country | Yes | ISO 3166-1 alpha-2 |
| docsUrl | No | Links to tender documents; omitted on the Free plan |
| subject | No | What is procured: lot names from the tender (for Armenian tenders, from the invitation document), in the original language (e.g. Armenian), separated by "; ". Omitted when unknown or the same as the title. |
| cpvCodes | No | Omitted on the Free plan |
| currency | Yes | ISO 4217 |
| deadline | Yes | Submission deadline, ISO 8601 UTC |
| lotsCount | Yes | |
| updatedAt | Yes | |
| budgetNote | No | Why budget is null; present only when it is |
| okpd2Codes | No | OKPD2 codes (EAEU sources); omitted on the Free plan |
| publishedAt | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint, idempotentHint and non-destructive, so safety is covered; the description adds genuinely non-derivable behaviour with the Free-plan gating that silently omits CPV codes, lots and document links. That caveat is exactly the kind of plan-dependent degradation an agent cannot infer from the schema or annotations. It loses a point only for not noting failure behaviour on an invalid or unknown id.
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 short sentences with zero filler: the payload contents come first, the id sourcing second, the plan limitation last. Each sentence carries distinct information and nothing repeats the schema or annotations verbatim.
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 single-parameter read tool with a full annotation set and an output schema, this is complete: it says what comes back, where the id comes from, and how results degrade on the Free plan. An agent has everything needed to call it correctly and to interpret missing fields.
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?
There is a single parameter with 100% schema description coverage, and the schema already says "Tender id from search_tenders results" — the description's instruction is essentially the same information. With the schema doing all the work on a one-param tool, the baseline of 3 applies; the description adds no format, length or validation detail beyond it.
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 ("Get the full card of one tender") and then enumerates exactly what the card contains: status, deadline, budget, subject, lots with prices and CPV codes, and document links. That enumeration lets an agent distinguish it immediately from search_tenders, which returns list-level results rather than a full card.
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?
"Use the id from search_tenders results" establishes the retrieval workflow and tells the agent where the required parameter comes from, effectively positioning this as the follow-up to a search. It stops short of explicit exclusions — nothing says when not to call it, e.g. whether a bare list view suffices — so it is clear context rather than a full when/when-not rule set.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Search open public procurement tenders (government tenders, госзакупки, тендеры) from official portals: EU TED and Armenia ARMEPS (armeps.am, gnumner.am). Free text in English, Russian or Armenian, optional filters: country, CPV/OKPD2 codes, budget, deadline. Returns up to 20 short cards; "subject" lists the lot names when the title alone does not say what is procured. Words are matched in the title, lot names and buyer. Budgets are in each tender's own currency (Armenia: AMD, from the invitation document); to compare budgets across countries, set budgetCurrency (e.g. "EUR") with minBudget/maxBudget.
| Name | Required | Description | Default |
|---|---|---|---|
| q | Yes | Free-text query in English, Russian or Armenian, e.g. "IT services, deadline > week" | |
| cpv | No | CPV codes or prefixes (e.g. ["72", "48"]) | |
| limit | No | Number of results, 1–20 (default 10) | |
| countries | No | Buyer countries, ISO 3166-1 alpha-2 (e.g. ["DE", "AM"]) | |
| maxBudget | No | Maximum estimated budget, in the tender currency | |
| minBudget | No | Minimum estimated budget, in the tender currency | |
| deadlineAfter | No | Only tenders whose deadline is after this ISO 8601 date/datetime | |
| budgetCurrency | No | Currency of minBudget/maxBudget (ISO 4217). When set, each tender budget is converted at fixed reference rates (updated quarterly), so EU (EUR, PLN, SEK…), Armenian (AMD) and Russian (RUB) tenders compare fairly; tenders in other currencies are skipped. When omitted, budgets are compared in each tender's own currency. |
| Name | Required | Description |
|---|---|---|
| hint | No | How to broaden the query when there are few or no results |
| results | Yes | |
| budgetNote | No | Why some results have budget null (not published or not yet read); present only when some do |
| degradedSources | No | Sources currently failing; data may be stale |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Adds substantial context beyond the annotations: result cap of 20 short cards, that matching covers title, lot names and buyer, that "subject" disambiguates what is procured, that budgets stay in each tender's native currency (AMD taken from the invitation document), and that conversion uses fixed quarterly reference rates while non-covered currencies are skipped. That is real invocation-shaping behavior, not a restatement of readOnly/idempotent hints.
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 what the tool is and where it searches, then filters, then output shape, then the budget-currency caveat. Dense but every sentence carries operational information; the multilingual keyword list (government tenders, госзакупки, тендеры) is slightly redundant padding.
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 an output schema present, return values needn't be spelled out, yet the description still flags the card-level granularity and the "subject" field, which is exactly the context needed to interpret results. Sources, currency behavior, and filter scope are all covered for an 8-parameter tool.
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 description earns above it by explaining the currency coupling between minBudget/maxBudget and budgetCurrency and warning that native-currency budgets are not directly comparable across countries. It adds less for the other parameters (q, cpv, countries, limit), which the schema already fully documents.
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 ("Search open public procurement tenders") and names the covered sources (EU TED, Armenia ARMEPS) plus multilingual query support. It implicitly distinguishes itself from get_tender by promising "up to 20 short cards," but never names the sibling or states that detail lookup is a separate call.
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 clear usage context: free text in English/Russian/Armenian, optional filters for country, CPV/OKPD2, budget and deadline, and an explicit "to compare budgets across countries, set budgetCurrency" instruction. It stops short of stating when not to use it or pointing to get_tender for full record details.
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.
add_to_watchlist2 fields changedInput schema / properties / channels / descriptionAdded value: +"Where alerts go: email (default) and/or an HTTPS webhook; webhooks need the Pro plan, Free gets a daily email digest"Input schema / properties / filters / descriptionAdded value: +"Which tenders to watch: the same query and filters as search_tenders (q, countries, cpv, budget, deadlineAfter); filters that match every tender are rejected"add_to_watchlistget_deadline_alertsget_tendersearch_tendersEU public tenders for AI agents: open tenders, award history, competitors and bid intelligence.
Government tender search for AI agents. UK, EU and US procurement opportunities.
Search EU public tenders across TED and 8 national portals. Monitor, match, and analyse procurement.
Polish public procurement for AI agents: BZP and TED tenders, awards, buyers and contractors.
Add one secure layer between your agents and this server.
