Servidor MCP para el actor LinkedIn Post Tracker and Comment Capture de Mamba Labs en Apify.
Apúntalo a perfiles de personas o páginas de empresa de LinkedIn. Devuelve sus publicaciones recientes como filas planas, con los recuentos reales de reacciones y comentarios en cada una, además de los comentaristas que LinkedIn muestra públicamente. Sin cookies, sin cuenta de LinkedIn, sin credenciales de ningún tipo.
npx -y @mambalabsdev/mcp-linkedin-post-engager-capture{
"mcpServers": {
"mamba-linkedin-post-engager-capture": {
"command": "npx",
"args": ["-y", "@mambalabsdev/mcp-linkedin-post-engager-capture"],
"env": { "APIFY_TOKEN": "your-apify-token" }
}
}
}Obtén un token de Apify en console.apify.com/account/integrations.
Related MCP server: LinkedIn Data API MCP Server
capture_linkedin_posts_and_commentersPerfiles de LinkedIn y páginas de empresa de entrada, sus publicaciones recientes y comentaristas públicos de salida.
Entrada | Tipo | Requerido | Notas |
| array | no | Perfiles de personas, por ejemplo |
| array | no | Páginas de empresa, por ejemplo |
| string | no | Fecha ISO, por ejemplo 2026-08-01. Las publicaciones anteriores a esta se omiten antes de que se cobre nada. |
| boolean | no | Una fila por comentarista público. Desactívalo para recopilar solo publicaciones. Por defecto |
| boolean | no | No devuelve filas de reacciones sin importar lo que establezcas, porque LinkedIn no muestra identidades de reaccionadores a un visitante sin sesión. Presente para que el límite sea visible en lugar de silencioso. Por defecto |
| integer | no | De 0 a 100. Diez es el propio límite de LinkedIn para un visitante sin sesión, así que subirlo por encima de 10 no hace nada. Ponlo a 0 para pagar solo por publicaciones. Por defecto |
| boolean | no | Desactivado por defecto, que es lo que asume el precio. Apify factura el ancho de banda residencial además de los eventos de este actor. Por defecto |
Nada es obligatorio. Una llamada sin ninguna de las listas de URL devuelve una fila de aviso no_input en lugar de un error, lo que refleja exactamente el actor.
Un conjunto de datos contiene tres tipos de fila, diferenciados por row_type. Filtra por él antes de cargar una tabla.
post es una fila por publicación de LinkedIn.
engager es una fila por comentarista público, unida a la publicación por post_id.
notice es un mensaje a nivel de ejecución o de fuente.
post_id es la parte numérica del URN de actividad de LinkedIn. Es estable entre ejecuciones, entre subdominios de país y entre ambas grafías de enlace permanente, por lo que es seguro como clave primaria y seguro como comprobación de "¿ya he publicado esto en Slack?".
Los límites de comentaristas son reales y no son pequeños. LinkedIn muestra unos diez comentarios de nivel superior a un visitante sin sesión, sea cual sea el total real. La cobertura medida en toda la ejecución fue de 178 comentaristas de los 4.857 que existían, lo que supone un 3,7 por ciento. Alrededor del 30 por ciento de las filas de comentarios no llevan marca de tiempo. Cada fila de publicación lleva commenters_collected junto a commenters_available, de modo que la cobertura en cualquier publicación dada es un número que puedes leer en lugar de algo que infieres.
Las identidades de los reaccionadores no están disponibles en absoluto para un visitante sin sesión. Cada fila de publicación sigue llevando el reaction_count real, y reactors_status dice unavailable_without_login.
Cada fila lleva degraded y degradation_reason. false significa que el actor miró; null en un campo significa que no pudo. Filtra por degraded antes de confiar en una ausencia. Dos cosas que parecen fallos no se degradan deliberadamente, porque ambas son mediciones: reactors_unavailable es cómo funciona LinkedIn sin sesión, y no_posts_found en una página que se cargó es una ausencia real.
Se te cobra por publicación recopilada y por comentarista recopilado, más una pequeña tarifa de inicio del actor. Las filas de aviso son gratuitas. posted_since filtra antes de que se cobre nada, por lo que una ejecución programada que no encuentre nada nuevo cuesta el inicio del actor y nada más.
No hay caché de resultados: cada ejecución vuelve a obtener, que es el comportamiento correcto para un actor cuyo trabajo es notar nuevas publicaciones. El precio está en la página de Apify del actor. Ejecutar este servidor consume créditos de Apify.
Es un cliente ligero para el actor de Apify. Pasa tu entrada y devuelve la salida del actor sin cambios. Cada comportamiento descrito anteriormente vive en el actor, no aquí.
Lee solo lo que LinkedIn sirve a un visitante sin sesión. No mantiene ninguna cookie de sesión, no usa ninguna cuenta de LinkedIn y no lee nada detrás de un inicio de sesión.
Las filas de comentaristas son personas con nombre. Cuando ejecutas esto, eres el responsable del tratamiento de los datos personales que devuelve y Apify es el encargado del tratamiento. Tu base legal es tuya para establecer; nada aquí es asesoramiento legal.
Este actor es no oficial y no está afiliado, patrocinado ni respaldado por LinkedIn o Microsoft.
Los errores se muestran, nunca se tragan. Una entrada no válida, un token no válido, un saldo agotado, un tiempo de espera o una ejecución que devuelva algo que no sea un conjunto de datos, todo vuelve como un error de herramienta explícito en lugar de un resultado vacío.
El actor está en la tienda de Apify. Este envoltorio tiene licencia MIT.
Construido por Mamba Labs.
Point it at LinkedIn person profiles or company pages and it returns the posts a search engine has indexed for each one, as flat rows: permalink, author, a text excerpt, and the publish date decoded from the post id. No request is sent to LinkedIn and no cookie or account is used. A result counts only when its permalink names the source, so posts are never attributed to the wrong author. Coverage is what the search index holds: posts appear days or weeks after publishing and many never appear. Commenters, reaction counts, comment counts and media are no longer returned (removed 2026-10-12, they came only from LinkedIn pages). Rows carry row_type post or notice; filter on row_type before loading a table. post_id is the activity URN and is stable across runs. Every row carries degraded and degradation_reason. Requires an APIFY_TOKEN and consumes Apify credits. Read only.
| Name | Required | Description | Default |
|---|---|---|---|
| company_urls | No | LinkedIn company page URLs, for example https://www.linkedin.com/company/microsoft. Supply this, profile_urls, or both. | |
| posted_since | No | ISO date or timestamp, for example 2026-08-01. Posts published before this are skipped BEFORE anything is charged. Leave empty to take everything the index holds. | |
| profile_urls | No | LinkedIn person profile URLs, for example https://www.linkedin.com/in/williamhgates. Country subdomains such as uk.linkedin.com are fine. Supply this, company_urls, or both. | |
| max_charge_usd | No | Optional. The actor's own spend cap for this run, in USD. The run stops before a charge would take its total past this amount, writes one notice row saying so, and keeps every row already charged. Leave it empty for no cap of your own; the maximum cost per run set in Apify is honored either way. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Goes well beyond the annotations: no request is sent to LinkedIn, no cookie or account is used, results only count when the permalink names the source, coverage is bounded by the search index, and commenters/reactions/media were removed on 2026-10-12. It also discloses the APIFY_TOKEN requirement, credit consumption, degraded/degradation_reason flags, and that post_id is a stable activity URN.
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 behavior, then layered with caveats, each of which carries real decision value (coverage limits, row_type, removed features, credits). It is long, but nearly every sentence prevents a specific failure mode; only minor tightening is possible.
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 fully describes the return shape (permalink, author, text excerpt, decoded publish date, row_type, post_id, degraded flags). Combined with the cost, auth, and coverage disclosures, an agent has everything needed to invoke and interpret this tool correctly.
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 100%, so all four parameters (company_urls, profile_urls, posted_since, max_charge_usd) are already fully documented in the schema, including the charge-cap behavior. The description adds no parameter-level detail beyond that, which is the correct baseline when the schema does the heavy lifting.
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 — it returns search-indexed LinkedIn posts for supplied person profiles or company pages as flat rows — and names the exact fields returned. It also proactively corrects the misleading tool name by stating that commenters are no longer returned. An agent knows precisely what comes back 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 concrete operating guidance: filter on row_type before loading a table, use posted_since to avoid charges, and expect posts to appear days or weeks after publishing with many never indexed. There are no sibling tools to route against, so explicit alternatives cannot be named, but the coverage caveats make the appropriate context clear.
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.
v1.4.0capture_linkedin_posts_and_commenters1 field changedInput schema / properties / max_charge_usdAdded value: +{
+ "description": "Optional. The actor's own spend cap for this run, in USD. The run stops before a charge would take its total past this amount, writes one notice row saying so, and keeps every row already charged. Leave it empty for no cap of your own; the maximum cost per run set in Apify is honored either way.",
+ "minimum": 0,
+ "type": "number"
+}v1.3.0capture_linkedin_posts_and_commenters8 fields changedInput schema / additionalPropertiesRemoved value: -falseInput schema / properties / collect_commentersRemoved value: -{
- "description": "Return one row per person who commented, with name, profile URL, comment text and comment likes. Turn it off to collect posts only, in which case commenters_collected comes back null rather than 0, because the actor did not look. Default: true.",
- "type": "boolean"
-}Input schema / properties / collect_reactorsRemoved value: -{
- "description": "LinkedIn serves no reactor identities to a logged-out visitor, so this returns no reaction rows whatever you set. It exists so the limit is visible rather than silent, and it adds one notice row per run. Set it to false to suppress that row. Default: true.",
- "type": "boolean"
-}Input schema / properties / company_urls / descriptionPrevious value: -"LinkedIn company page URLs whose recent posts you want, for example https://www.linkedin.com/company/microsoft. Supply this, profile_urls, or both."New value: +"LinkedIn company page URLs, for example https://www.linkedin.com/company/microsoft. Supply this, profile_urls, or both."Input schema / properties / max_engagers_per_postRemoved value: -{
- "description": "Caps the engager rows charged per post. LinkedIn shows about ten comments to a logged-out visitor, so ten is the platform ceiling and raising this above ten does nothing. Set it to 0 to pay for posts only. Default: 10.",
- "maximum": 100,
- "minimum": 0,
- "type": "integer"
-}Input schema / properties / posted_since / descriptionPrevious value: -"ISO date or timestamp, for example 2026-08-01. Posts published before this are skipped BEFORE anything is charged, so a scheduled run that finds nothing new costs the actor start and nothing else. Leave empty to take everything the page advertises."New value: +"ISO date or timestamp, for example 2026-08-01. Posts published before this are skipped BEFORE anything is charged. Leave empty to take everything the index holds."Input schema / properties / profile_urls / descriptionPrevious value: -"LinkedIn person profile URLs whose recent posts you want, for example https://www.linkedin.com/in/williamhgates. Country subdomains such as uk.linkedin.com are fine. Supply this, company_urls, or both."New value: +"LinkedIn person profile URLs, for example https://www.linkedin.com/in/williamhgates. Country subdomains such as uk.linkedin.com are fine. Supply this, company_urls, or both."Input schema / properties / use_residential_proxyRemoved value: -{
- "description": "Off by default, which is what the pricing assumes and what was measured. Turn it on only if LinkedIn starts refusing the platform's datacenter addresses. Apify bills residential bandwidth on top of this actor's events. Default: false.",
- "type": "boolean"
-}v1.0.1capture_linkedin_posts_and_commentersScored across 1 tool
There is only one tool, so there is no possibility of selecting the wrong one. Its purpose (capture indexed LinkedIn posts for given profiles/pages) is stated unambiguously.
The single name uses a clear snake_case verb_noun form, which is internally consistent. However, it still advertises 'commenters' even though the description says commenter data was removed, making the name stale and slightly misleading.
The server exposes exactly one tool for a very narrow, read-only capture operation, so the count is defensible but thin. An agent gets no separate means to list sources, filter row types, or fetch metadata without parsing the single flat output.
For a read-only 'capture posts from a search index' domain, one tool covers the whole lifecycle and returns stable post_id, row_type, and degradation flags. Gaps are minor: no dedicated source-validation or post-filtering operation, and the name's promise of commenters is no longer fulfilled.
Scrape and analyze public LinkedIn posts as structured JSON via the Apify LinkedIn Posts API.
LinkedIn companies, hiring signals and public profiles by URL. No login, no cookies. Pay per result.
Live LinkedIn data for AI agents: profiles, companies, jobs, posts, email finding. No account risk.
Search LinkedIn & Sales Navigator. Get structured people, companies, posts, jobs data. 4.9/5 on G2.