IAdetector
NUGET · IADETECTOR.MCP
Detecta texto generado por IA, 100% local, sin enviar tu texto a ningún servidor.
Available components
Last scanned (17 hours ago)
Scores higher than 50% of NuGet servers
Category AI and Machine Learning
How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. How we score → Why this is hard to score →
Supply Chain Security100
- No malware found by supply-chain analysis.Pass
- No known CVEs affecting this package version or its production dependencies.Pass
- No install/post-install scripts declared.Pass
- No production dependencies, so there is no dependency health to assess. View diagnostics → Pass
Provenance & Transparency48
- Source repository is publicly reachable at the declared URL. View diagnostics → Pass
- Build provenance not yet verified. View diagnostics → Unverified
- Clear OSI-approved license (MIT).Pass
- Actively maintained (last published 0 days ago).Pass
- Publishes a security disclosure policy (SECURITY.md).Pass
Schema Quality & AI Usability72
- AI-judged instruction clarity (excellent).Pass
- Context-footprint check failed: tool/resource definitions use about 949 tokens (~189/item across 5 items; 5 tools + 0 resources), over budget; trim descriptions and params. See how to fix → Fail
- Usage-examples check failed: none of the tools include examples. See how to fix → Fail
Stability & Change Management33
- Stability observed for 10 of 30 days with no destabilising changes; credit accrues until the full window elapses.Partial
Tool Coverage100
- 100% of tools have a non-trivial description (not blank, and not just the tool's name).Pass
- 100% of tool parameters carry a description.Pass
Tool Safety100
- No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.Pass
- We read all 5 captured tool definition(s), and no name or description among them implies an irreversible operation.Pass
- An AI judge read all 5 captured unit(s) of tool text and found none that tries to manipulate the model reading it.Pass
Capabilities100
- Implements a supported MCP spec version (2025-11-25); the latest is 2026-07-28.Pass
Checks we ran (15)
- Malware scan
- Known CVEs
- Install scripts
- Dependency health
- Source repository
- License
- Maintenance
- Security disclosure
- Schema quality
- Stability
- Tool coverage
- Injection markers
- Destructive annotations
- Judged manipulation
- MCP protocol
Checks we could not complete (1)
These are gaps on our side, not findings about this package. We only credit what we can confirm, so a category made up entirely of them scores 0.
- Provenance in Provenance & TransparencyBuild provenance not yet verified.
Does not apply to this package (2)
- Endpoint SecurityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
- Transport & ReachabilityScored on hosted endpoints only. A package has no live endpoint of its own for us to probe.
The score does not measure everything. What the score does not measure →
Every change we have recorded for this component, newest first. Security-relevant changes are always shown. ▲ marks a change for the better, ▼ a change for the worse; unmarked changes are neutral.
- 21 Nov 26 +4
- Stability: unverified → 0.33 ▲ functional
- Package version: 1.0.1 → 1.1.0 functional
- 12 Nov 26 +15
- Malware scan: unverified → pass ▲ security
- 11 Nov 26 56
First indexed and scored.
Diagnostic detail from the automated scan of this channel: what the scanner observed at each step, so you can see exactly where a check passed or failed. It is informational only and never changes the trust score.
Captured 21 Nov 2026 · Analysed nuget/IAdetector.Mcp@1.1.0
Provenance Inconclusive
We could not complete the check, so nothing is claimed either way. This is a gap on our side, not a finding about the package.
| Result | Inconclusive |
|---|---|
| Ecosystem | nuget |
| Reason | Signature present, unverified |
Signer
| Signed as | Repository signature, names nuget.org |
|---|---|
| Subject | CN=NuGet.org Repository by Microsoft,O=NuGet.org Repository by Microsoft,L=Redmond,ST=Washington,C=US |
| Issuer | CN=DigiCert Trusted G4 Code Signing RSA4096 SHA384 2021 CA1,O=DigiCert\, Inc.,C=US |
| Valid from | 23 Feb 2024 |
| Valid until | 18 May 2027 |
| Service index | https://api.nuget.org/v3/index.json |
| Owners | JOSETRA44 |
Background: How many MCP packages publish verified provenance →
Dependencies 0 packages
| Packages resolved | 0 |
|---|---|
| Tree resolution | Complete |
Background: SBOMs and build attestations, explained →
Registry declarations For information only
What the registry and the package's own manifest say about this version. The publisher or the registry writes these, nothing here is verified, and none of it affects the score.
| Repository type | git |
|---|---|
| Commit | 7034c775b5de87688784a180c84f3495d56b16de |
| Project URL | https://github.com/JOSETRA44/IAdetector-mcp |
The tools this component advertises to a client, with an estimated token cost for each. Expand a tool to see its parameters and schema. The per-tool counts are indicative and are not scored directly; the schema's total context footprint is one signal in Schema Quality & AI Usability. A tool's description is untrusted text the model reads on every call, which is what makes this list a security surface and not just an inventory: how tool poisoning works →
batch_detect_directory ~147
Analiza todos los archivos de un directorio que coincidan con un patrón (por defecto *.txt), devolviendo un array JSON con un resultado (o error) por archivo. Úsala en vez de llamar a detect_file repetidamente cuando hay más de un archivo que analizar.
| Name | Type | Req | Description |
|---|---|---|---|
| maxFiles | integer | – | Número máximo de archivos a procesar en esta llamada (entre 1 y 1000; por defecto 100). |
| path | string | yes | Ruta absoluta del directorio a analizar. |
| pattern | string | – | Patrón glob de archivos a incluir, ej. '*.txt'. |
| recursive | boolean | – | Si es true, busca también en subdirectorios. |
No output schema declared.
No examples provided.
compare_versions ~202
Compara un texto original con su reescritura (sin guardar estado: pasa ambas en cada llamada). Devuelve las dos detecciones, deltaProbability, el cambio por señal con su contribución al logit, y outcome: mejora / empeora / sin_cambio / no_concluyente. Salvaguarda anti-optimización-vacía: la similitud léxica de contenido (palabras de ≥4 letras compartidas) debe ser ≥ 0,5; si la reescritura pierde vocabulario del original, el outcome es no_concluyente aunque la probabilidad baje, porque el texto puede haber perdido su significado. Úsala en cada iteración de un bucle de reescritura para decidir el siguiente paso con explain_detection.
| Name | Type | Req | Description |
|---|---|---|---|
| originalText | string | yes | Texto original (máximo 20000 caracteres). |
| rewrittenText | string | yes | Texto reescrito (máximo 20000 caracteres). |
No output schema declared.
No examples provided.
detect_file ~135
Analiza el contenido de un archivo de texto local (máximo 5MB) y estima la probabilidad de que haya sido generado por IA; fragmenta y, si es muy largo, muestrea automáticamente el documento. Devuelve el mismo JSON que detect_text (veredicto, probabilidad, confianza, señales). Para texto pegado directamente en la conversación usa detect_text en su lugar; para varios archivos a la vez usa batch_detect_directory.
| Name | Type | Req | Description |
|---|---|---|---|
| path | string | yes | Ruta absoluta del archivo de texto a analizar. |
| verbose | boolean | – | Si es true, incluye el desglose completo de cada señal individual. |
No output schema declared.
No examples provided.
detect_text ~263
Analiza un texto pegado directamente (no un archivo) y estima la probabilidad de que haya sido generado por IA, combinando 7 señales estadísticas y 2 modelos ONNX locales (sin llamar a ninguna API externa). Devuelve un JSON con veredicto, probabilidad de IA, confianza y desglose de señales. Para el contenido de un archivo en disco usa detect_file en su lugar — no lo leas tú mismo para pegarlo aquí. El texto analizado NUNCA sale de la máquina. El resultado se cachea con clave (texto + versión de los modelos), así que reanalizar el mismo texto es instantáneo. AL COMUNICAR EL RESULTADO: es evidencia probabilística, nunca prueba. Los falsos positivos afectan de forma desproporcionada a quien escribe en un idioma que no es el suyo y a la prosa técnica formulaica, y un texto de menos de un párrafo da confianza baja y poco informativa. No lo presentes como base suficiente para acusar a nadie.
| Name | Type | Req | Description |
|---|---|---|---|
| text | string | yes | Texto a analizar (máximo 20000 caracteres; para textos más largos usar detect_file). |
| verbose | boolean | – | Si es true, incluye el desglose completo de cada señal individual con su explicación. |
No output schema declared.
No examples provided.
explain_detection ~202
Explicación PRESCRIPTIVA de una detección, pensada para iterar reescribiendo: por cada señal devuelve si midió (hasData), su puntuación, cuánto empuja el resultado (logitContribution = coeficiente calibrado × puntuación; ordena por ahí para saber qué atacar primero), el rango típico en texto humano de referencia (percentiles 10/50/90 del corpus de calibración) con su estado (en rango / más de IA que el rango humano / sin datos) y, cuando está fuera de rango, qué cambiar en la reescritura. Incluye prioritizedActions con las señales fuera de rango ordenadas por contribución. Los rangos son del corpus HC3 (inglés): una referencia para este detector, no una verdad universal. Para comparar una reescritura ya hecha con su original usa compare_versions.
| Name | Type | Req | Description |
|---|---|---|---|
| text | string | yes | Texto a analizar (máximo 20000 caracteres). |
No output schema declared.
No examples provided.
How do I install the IAdetector MCP server?
IAdetector runs locally as a NuGet package, launched with dnx IAdetector.Mcp@1.1.0 --yes. Ready-made configuration for Claude, Cursor, VS Code, Codex and 3 more is on this page, copied from each client's own documentation.
nuget · IAdetector.Mcp
claude mcp add josetra44-iadetector-mcp -- dnx IAdetector.Mcp@1.1.0 --yes
What is the IAdetector MCP server?
IAdetector is an MCP server listed in the public MCP registry as io.github.JOSETRA44/iadetector-mcp. Detecta texto generado por IA, 100% local, sin enviar tu texto a ningún servidor. This page covers its NuGet package (IAdetector.Mcp).
Is the IAdetector MCP server safe to use?
IAdetector scores 75 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 21 November 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.
What tools does the IAdetector MCP server expose?
IAdetector exposes 5 tools: detect_text, batch_detect_directory, explain_detection, detect_file, compare_versions. Their descriptions and schemas cost roughly 949 tokens of context every time the server is loaded.
Is the IAdetector MCP server still maintained?
IAdetector is still listed as active in the MCP registry. We last reached this channel on 21 November 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.
What licence is the IAdetector MCP server under?
IAdetector declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.
What did VerifyMCP test on the IAdetector MCP server?
VerifyMCP ran 15 checks on IAdetector, covering Malware scan, Known CVEs, Install scripts, Dependency health and 11 more. One item could not be completed, a gap on our side rather than a finding about IAdetector: Provenance. Endpoint Security and Transport & Reachability do not apply to a NuGet package, so they were not tested. The score does not measure everything, and the scoring documentation lists what it leaves out.