Research for Monarchic products.

ExplicitMem is the first Monarchic product with published benchmark results. Available MCP pages document the request, output, access required, and current limits.

Read ExplicitMem research Browse products

Published research / 1

ExplicitMem MCP

Store named memory objects and retrieve them in later agent sessions without relying on chat history that disappears between sessions.

Product / Agent Memory

Available

Memory that survives long agent work

Teams running agents over long projects where context must survive sessions.

Write memorySearch memoryRetrieve contextAudit objects

Open product details

Published benchmark / LongMemEval-S

99.8%

Answer accuracy / 500 questions

Faithfulness: 100%
Recall@K: 100%

Four evaluation views cover LongMemEval-S synthesis, LoCoMo retrieval, cross-dataset candidate generation, and generic-answer support.

Each result applies only to its named dataset, configuration, metric, and artifact. The studies do not establish a cross-product ranking, general agent accuracy, or hosted latency guarantee.

Method and results Use case and access

Available products

Read the product contract

Every available MCP page documents a concrete call, output shape, required access, current limits, and runtime usage category.

RepoIntel MCP

Repository Intelligence

Workflow and limits

Give agents repository maps, generated documentation, workflow context, and source-file citations.

IncidentOps MCP

Incident Response

Workflow and limits

Combine alerts, deploys, logs, ownership, and response notes into a coherent incident picture.

InfraProfiler MCP

Infrastructure Intelligence

Workflow and limits

Connect infrastructure telemetry and pipeline data, then rank the bottlenecks that matter most.

See all 17 available MCPs, including their calls, outputs, access requirements, and current limits.

Open hosted MCP catalog

Publication policy

What we publish

Product availability is assessed separately; catalog entries do not need uniform research coverage. Each publication stays attached to the product, method, date, and scope it actually evaluates.

External benchmarks

Runs against published datasets. We include the date, method, result, and limits of the comparison.

Read benchmark results

Monarchic evaluations

Tests designed by Monarchic for a specific capability, failure mode, dataset, or release condition.

Read test methods

Engineering checks

Automated tests, security checks, isolation tests, protocol checks, and production measurements.

Read security details