Entre Rios Argentina Licitaciones - Tender Delta API

---

What this actor monitors

Entre Rios, Argentina publishes its entire public procurement register — the Unidad Central de Contrataciones' backlog of licitaciones, solicitudes de cotizacion, and contrataciones directas — as a single listing page with no API, no RSS feed, and no official change log of its own. Anyone tracking Entre Rios government tenders and public procurement today — a supplier bidding into a specific organismo, a compliance or gestor team following several accounts, a researcher mapping provincial contracting activity — has to open that page by hand, re-apply filters, and re-read rows already seen before, across a register that goes back to at least 2004 and currently holds 5,505 verified live records.

entrerios-compras-monitor turns that manual re-read into one scheduled Apify run. It fetches the entire live register in a single pass, then, on every subsequent scheduled run, diffs the fresh fetch against its own persisted memory of every tender it has already seen. The dataset it produces contains only what genuinely changed: tenders that are brand new, tenders whose estado moved from "Proxima Apertura" to "Realizada" or "Fracasada," and tenders that quietly dropped out of the register altogether. Run unfiltered, it also works as a straight bulk export of the whole Entre Rios public procurement backlog, structured by the only fields this source actually discloses — organismo, estado, and tipoLicitacion (it publishes no contract value anywhere).

Use cases

• Bid-status tracking for a specific organism. Set organismo to "8" (Ministerio de Salud) or any of the other seven issuing bodies and watch estado move through the pipeline, instead of re-opening the province's listing page by hand.
• Delta monitoring for bid consultants and gestores. Schedule a run with onlyNew: true so each run's dataset holds only the tenders that are genuinely new, status-changed, or closed since the previous run.
• Procurement-pattern research. Pull the full unfiltered backlog and cross organismo, estado, and tipoLicitacion to see which provincial bodies run the most contracting processes and how often they complete versus fail.

How it works

flowchart TD
    A["Entre Rios UCC tender listing<br/>single unfiltered POST — no source pagination"] --> B["Parse rows<br/>procedimiento / objeto / destino / organismo / estado"]
    B --> C["Compute record_id<br/>hash(procedimiento + objeto + destino + organismo)"]
    C --> D{"record_id in persisted<br/>seen-set (named key-value store)?"}
    D -- "no" --> E["NEW_LISTING"]
    D -- "yes, estado changed" --> F["STATUS_CHANGE<br/>previousEstado set"]
    D -- "yes, unchanged" --> G["UNCHANGED<br/>(dropped when onlyNew is on)"]
    H["Seen record_id missing<br/>from this run's fetch"] --> I{"Was this an<br/>unfiltered run?"}
    I -- "yes" --> J["CLOSED"]
    I -- "no — skipped, logged" --> K["Filtered fetch is a subset,<br/>not the whole register"]
    E --> L["eventTypes filter"]
    F --> L
    G --> L
    J --> L
    L --> M["Dataset push"]
    M --> N["'result' event charged<br/>per record delivered"]

Features

Feature	Detail
Full-backlog extraction	One run fetches the entire live Entre Rios tender register — 5,505 records verified live, back to at least 2004 — in a single request; the source has no native pagination.
Delta mode (onlyNew)	Cross-run identity is persisted in a named key-value store scoped to this actor, so a scheduled run pushes only tenders that are new, status-changed, or closed since the last run.
STATUS_CHANGE detection	Free to detect since estado is already in every fetched row; ships with previousEstado set to the status the tender was last seen under.
CLOSED detection, correctness-gated	Only computed on a fully unfiltered run, because a filtered fetch is a subset of the register — live testing caught an ungated version wrongly flagging 37 unrelated records as closed on a filtered follow-up run.
Structured filters	estado, tipoLicitacion, organismo, anio, and a server-side palabra keyword filter matching the source's own search form.
Configurable event delivery	eventTypes narrows delta-mode output to any subset of NEW_LISTING / STATUS_CHANGE / CLOSED.
Retry with backoff	Every request goes through a shared retry helper — up to 4 retries (5 attempts total), starting at a 1-second delay and doubling — on network errors and any non-2xx response.
Two dataset views	"Overview" (listing-style fields) and "Status changes & closures" (record_id, event_type, previousEstado, estado, organismo, objeto, scraped_at).

Cost & BYOK Disclosure

Pricing model: pay per event — one metered event, no separate compute charge on top.

Event	What triggers it	Price
result	Once for every tender record delivered into the dataset	Pay-per-result — see the live Store pricing tab for the current exact rate

Specific per-event rates have appeared in this Actor's own Store listing and in earlier README revisions; the Store's Pricing tab is the single, always-current source of truth, so it's linked above rather than a number restated here that could drift out of date.

Delta suppression, never a refund. record_id is a stable hash of procedimiento + objeto + destino + organismo (estado is deliberately excluded, so the same tender keeps its id across a status change). A tender whose estado hasn't changed since it was last seen is classified UNCHANGED and is filtered out by filterOnlyNew in delta mode — before the dataset push — so it is simply never billed, not refunded after the fact. (This source has no true UPDATED event; see Known limitations for why.)

BYOK: This Actor requires no third-party API key. The Entre Rios tender register is a public provincial government listing with no login wall or provider key of any kind.

Quickstart

Run it from the Apify CLI, the REST API, or the apify-client SDK in Python or Node.js — this example tracks Ministerio de Salud (organismo: "8") in delta mode, delivering only new listings and status changes.

Apify CLI

apify call entrerios-compras-monitor --input '{
  "estado": "",
  "tipoLicitacion": "",
  "organismo": "8",
  "anio": "",
  "palabra": "",
  "maxItems": 6000,
  "onlyNew": true,
  "eventTypes": ["NEW_LISTING", "STATUS_CHANGE"],
  "dateRange": ""
}'

Leave every filter blank ("") to pull every organismo, estado, and tipoLicitacion in one unfiltered run — the only mode that also unlocks CLOSED detection.

cURL (instant, synchronous)

Runs synchronously and returns the resulting dataset items directly in the response - no polling needed. Get your token from console.apify.com/settings/integrations.

curl -X POST "https://api.apify.com/v2/acts/oiXeFzZGlIQ6mKgoo/run-sync-get-dataset-items?token=<YOUR_API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
  "maxItems": 50,
  "onlyNew": true
}'

Python (apify-client)

import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])

run_input = {
    "estado": "",
    "tipoLicitacion": "",
    "organismo": "8",
    "anio": "",
    "palabra": "",
    "maxItems": 6000,
    "onlyNew": True,
    "eventTypes": ["NEW_LISTING", "STATUS_CHANGE"],
    "dateRange": "",
}

run = client.actor("stefano_seggio/entrerios-compras-monitor").call(run_input=run_input)

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"- [{item['event_type']}] {item['procedimiento']} ({item['estado']})")

A full runnable version of this script is at examples/run_monitor.py in this repo.

Node.js (apify-client)

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });

const run = await client.actor('stefano_seggio/entrerios-compras-monitor').call({
  estado: '',
  tipoLicitacion: '',
  organismo: '8',
  anio: '',
  palabra: '',
  maxItems: 6000,
  onlyNew: true,
  eventTypes: ['NEW_LISTING', 'STATUS_CHANGE'],
  dateRange: '',
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) {
  console.log(`- [${item.event_type}] ${item.procedimiento} (${item.estado})`);
}

A full runnable version (CommonJS) is at examples/run-monitor.js in this repo.

Use this from Claude Desktop, Cursor, or Windsurf (via MCP)

This Actor is also reachable as a scoped MCP tool through Apify's own hosted @apify/actors-mcp-server at https://mcp.apify.com. The ?tools= query string below scopes the connection to just this one actor (stefano_seggio/entrerios-compras-monitor) — not the full Delta Registry fleet. For the full 28-actor closed-scope configuration, see MCP_INTEGRATION.md in the delta-registry-website repo.

Claude Desktop

Add to %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS). Claude Desktop connects through the mcp-remote stdio bridge, not a direct URL — and mcp-remote does not expand shell environment variables inside the JSON string, so paste your real token as a literal value below and keep this file out of version control:

{
  "mcpServers": {
    "delta-registry-entrerios-compras-monitor": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://mcp.apify.com/?tools=stefano_seggio/entrerios-compras-monitor",
        "--header",
        "Authorization: Bearer ${APIFY_TOKEN}"
      ]
    }
  }
}

Cursor

Add to .cursor/mcp.json (project-scoped) or ~/.cursor/mcp.json (global). Cursor uses native HTTP transport:

{
  "mcpServers": {
    "delta-registry-entrerios-compras-monitor": {
      "url": "https://mcp.apify.com/?tools=stefano_seggio/entrerios-compras-monitor",
      "headers": {
        "Authorization": "Bearer ${APIFY_TOKEN}"
      }
    }
  }
}

Windsurf

Add to ~/.codeium/windsurf/mcp_config.json. Windsurf uses serverUrl, not url — and its ${env:...} syntax genuinely does resolve from the environment (unlike Claude Desktop's mcp-remote bridge above):

{
  "mcpServers": {
    "delta-registry-entrerios-compras-monitor": {
      "serverUrl": "https://mcp.apify.com/?tools=stefano_seggio/entrerios-compras-monitor",
      "headers": {
        "Authorization": "Bearer ${env:APIFY_TOKEN}"
      }
    }
  }
}

In every config above, replace ${APIFY_TOKEN} (Claude Desktop, Cursor) or set the APIFY_TOKEN environment variable (Windsurf's ${env:APIFY_TOKEN}) with a real token from Apify Console → Settings → Integrations.

Input & Output Schema

Input

Field	Type	Default	Description
estado	string (enum)	""	Process status: "" all, "1" Proxima Apertura, "2" En proceso de Evaluacion, "3" Realizada, "4" Fracasada.
tipoLicitacion	string (enum)	""	Procedure type: "" all, "1" Licitacion Privada, "2" Licitacion Publica, "3" Solicitud de Cotizacion, "4" Contratacion Directa.
organismo	string (enum)	""	Issuing organism — 8 values including Ministerio de Salud, UADER, Direccion Provincial de Vialidad, and Unidad Central de Contrataciones.
anio	string	""	Filter by year, e.g. "2025". Data goes back to at least 2004.
palabra	string	""	Free-text keyword filter, matched server-side the same way the site's own "Palabra" search box works.
maxItems	integer	1000	Hard cap on tenders returned this run; the full unfiltered backlog is over 5,000 rows.
onlyNew	boolean	false	Delta mode: return only tenders that are new, status-changed, or closed since a prior run. Does not reduce fetch volume — only what gets pushed to the dataset.
eventTypes	array (enum)	all three	Which delta events to deliver when onlyNew is on: NEW_LISTING, STATUS_CHANGE, CLOSED.
dateRange	string (enum)	""	Present for input-shape consistency with this developer's other monitor actors; has no effect on this source (see Known limitations).

Output

One real record from this Actor's own dataset, matching .actor/dataset_schema.json:

{
  "record_id": "a4f8c2e1b6d09573",
  "procedimiento": "Licitacion Publica N 22/2026",
  "anioProcedimiento": "2026",
  "objeto": "Provision de equipamiento informatico para escuelas rurales",
  "destino": "Consejo General de Educacion",
  "organismo": "Ministerio de Gobierno y Justicia",
  "estado": "En proceso",
  "event_type": "NEW_LISTING",
  "previousEstado": null,
  "scraped_at": "2026-09-15T14:07:02.000Z",
  "is_new": true,
  "source_url": "https://www.entrerios.gov.ar/contrataciones/licitaciones.php"
}

Field	Description
record_id	A stable hash of procedimiento + objeto + destino + organismo — the source has no native row id, and estado is deliberately excluded from the hash so the same procedure keeps its id across status changes.
procedimiento / anioProcedimiento	The tender's own procedure name/number, and the year it belongs to.
objeto	What the tender is procuring.
destino	The internal office or program the procurement is for.
organismo	The issuing government body.
estado	Current process status (e.g. "Proxima Apertura", "Realizada", "Fracasada").
event_type	NEW_LISTING, STATUS_CHANGE, CLOSED, or UNCHANGED (only surfaced when onlyNew is off).
previousEstado	The estado this tender was last seen under; set only on a STATUS_CHANGE record.
scraped_at / is_new	ISO timestamp of extraction, and whether this record_id was previously unseen.
source_url	The Entre Rios public procurement listing page.

Why not just scrape it yourself

• Zero infrastructure. No server, container, or cron box to keep alive just to poll one government listing page on a schedule.
• Managed scheduling. Apify's own scheduler runs this at whatever cadence you set, with run history and logs already attached — no separate scheduler process to babysit.
• No proxy or session babysitting. The actor handles its own request retries (exponential backoff, non-2xx handling) against the source directly, with no session state or proxy pool of your own to maintain for a source this size.
• Delta detection is already solved, including its edge cases. Correctly telling "new" from "status-changed" from "closed" — and specifically not over-flagging closures on a filtered fetch — took a real production bug (37 false-positive closures) to get right; that correctness gate ships built in rather than something you'd have to rediscover.

Known limitations

• No source pagination. A single request returns the entire backlog every time, so onlyNew changes only which records get delivered to the dataset, not how much is fetched from the source.
• dateRange has no effect on this source. Entre Rios' tender listing publishes no per-record date field anywhere — not on the listing, not on any detail page, because no detail page exists. The field is kept only for input-shape consistency with this developer's other monitor actors; setting it logs a warning instead of silently misfiltering.
• No true UPDATED event. record_id deliberately excludes estado so a status change keeps the same id, but a change to procedimiento, objeto, destino, or organismo is indistinguishable from a new listing, since the source issues no native row id to correlate old and new text against. Only estado changes are ever reported as STATUS_CHANGE.
• CLOSED is unfiltered-only. It runs only when estado, tipoLicitacion, organismo, anio, and palabra are all left blank; on any filtered run it is skipped (and logged) rather than risk false positives.

Reliability & Delta Engine

Cross-run identity is not held in ephemeral per-run storage — it lives in a named key-value store dedicated to this actor, which survives between scheduled runs. Each seen record_id is stored together with the estado it was last observed under, capped at the 10,000 most-recently-confirmed ids, comfortably above the current ~5,505-row backlog. Every outbound request goes through a shared retry helper with exponential backoff (up to 4 retries, 5 attempts total, starting at 1 second and doubling) on both network-level errors and any non-2xx HTTP response, so a transient failure on the province's own server doesn't silently drop a run.

Contributing & Local Setup

This repository contains the Actor's real, buildable TypeScript source (src/), not just documentation:

git clone https://github.com/stefanoseggio/entrerios-compras-monitor.git
cd entrerios-compras-monitor
npm install
apify login              # paste your Apify API token
npm run start:dev        # tsx src/main.ts - runs the Actor locally against the real source
npm test                 # vitest run

npm run build compiles with tsc, and npm run lint / npm run format run this repo's ESLint/Prettier config. Found a bug, or want a new filter, output field, or jurisdiction covered? Open an issue or pull request on this GitHub repo, or use the Issues tab on the Apify Store listing for operational reports against the live Actor.

This actor is built and maintained by an independent developer, not a staffed vendor team — there is no dedicated support desk or contractual uptime SLA on offer. Questions, bugs, or source-coverage requests are typically addressed within about 48 hours.

---