nano SIEM
Agents & MCPInvestigator

Investigator

Supported in:Self-hosted

nano-investigator: an MCP server that turns a coding agent into a SOC analyst workstation over your nano data

Connect nano-investigator to search, triage alerts, investigate cases, hunt, and build parsers from your coding agent. The local MCP server runs over stdio and calls nano’s API and search service for each operation.

The server registers 91 tools, five MCP prompts, and MCP resources.

It is published to npm as @nano-rs/investigator-mcp-server. These pages describe version 0.1.16.

A typical session looks like a conversation against your data:

  • "how many alerts fired today?" runs get_alert_counts
  • "what do we know about 10.5.2.40?" runs get_entity_context plus a scoped search
  • "investigate alert 456" runs the investigate_alert prompt
  • "who are the riskiest entities right now?" runs get_risky_entities
  • "build a parser for these logs" runs the build_parser prompt

How it differs from nanodac

nano-investigator and nanodac sit on opposite sides of the detection lifecycle and run side by side as separate MCP servers, sharing the same environment variables.

nanodacnano-investigator
RoleDetection-as-code authoringSOC operations + parser authoring
WorkWrite, validate, and deploy detection rulesSearch, triage, investigate, hunt, build parsers
DetectionsReads and writesReads only
Default postureWrites to the SIEMRead-first; writes cases, notebooks, and parsers

Detection-rule editing stays in nanodac. nano-investigator only reads detections; for the authoring side, see Detection-as-Code.

Log-source parsers can be authored two ways, and neither uses nanodac's tools (nanodac is detection-only). The version-controlled path drafts a parser.yaml against the nano-rs/parsers repo and validates it with the vector vrl CLI, then ships by PR or UI import — see Authoring parsers. The interactive path, new here, authors and deploys a parser directly against a live nano deployment over MCP — see Build parsers.

The surface at a glance

AreaToolsAccess
Search and schema9Read, plus save_search and create_shared_search
Alerts3Read-only
Cases13Read, plus 8 write tools (merge_cases is destructive)
Notebooks9Read, plus 5 write tools
Dashboards7Read, plus create and update
Detections and MITRE5Read-only
Risk5Read, plus reset_entity_risk (destructive)
Enrichment2Read-only
Prevalence3Read-only
System4Read-only
Parsers23Read, plus author, publish and import
Onboarding4Credentials and pull-source transports
Recon4Estate census and hunt drafts

Most of the surface reads. The tools that write:

  • Cases: create, update, change status, assign, attach alerts, post wall entries, merge notebooks into cases, and merge cases.
  • Notebooks: create, append entries, add references, update and share.
  • Dashboards: create and update.
  • Risk: reset_entity_risk clears an entity's risk score.
  • Parsers: create, update, publish, deploy and undeploy log-source parsers, import from a parser repository and wire routing.
  • Onboarding: store credentials and create pull-source transports. import_credential_from_file reads a file from the machine the server runs on.
  • Search: save_search and create_shared_search.

Alerts can't be changed. There's no acknowledge, close or assign tool for alerts; triage happens by adding an alert to a case. Detections are read-only here; author them with nanodac.

Alongside the tools, the server registers:

  • 5 MCP prompts: investigate_alert, hunt_entity, hunt_campaign, morning_briefing and build_parser. These are workflows the agent follows, not callable tools.
  • MCP resources: an nPL reference, a VRL parser guide, seven investigation playbooks, and a UDM schema and SQL guide. See Tool Reference.

Troubleshoot

Need help with these checks? Contact support with the request ID and time with timezone.

A tool returns 403

The error names the endpoint. Add the matching permission from the tool reference to the key.

What's next

On this page