Grafana

purecipher-grafana · v0.4.0 · by purecipher
Access requirements not specifiedCertification: Basic
What you can do

7 product-specific SecureMCP tools to read Grafana data using your own connection. Select only the tools you need; provider permissions and profile approval are required.

Tools

7 tools

Explore the tools exposed by Grafana to MCP clients.

Registered SecureMCP definitions
7 of 7 tools
  • grafana_search_dashboards

    Search dashboards permitted by your Grafana service account.

    View inputs and schema
    limit
    Optional
    integer
    query
    Optional
    string
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "properties": {
        "limit": {
          "type": "integer",
          "default": 20
        },
        "query": {
          "type": "string",
          "default": ""
        }
      },
      "additionalProperties": false
    }
  • grafana_get_dashboard

    Read an accessible dashboard definition by UID; requires dashboards:read. Does not query its data sources.

    View inputs and schema
    uid
    Required
    string
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "required": [
        "uid"
      ],
      "properties": {
        "uid": {
          "type": "string"
        }
      },
      "additionalProperties": false
    }
  • grafana_list_datasources

    List accessible data-source identities and types; credentials, URLs and configuration are omitted. Grafana's endpoint has no pagination.

    View inputs and schema
    max_results
    Optional
    integer
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "properties": {
        "max_results": {
          "type": "integer",
          "default": 20
        }
      },
      "additionalProperties": false
    }
  • grafana_get_datasource

    Read one data-source's identity/type by UID; excludes credentials and configuration. Requires datasources:read.

    View inputs and schema
    uid
    Required
    string
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "required": [
        "uid"
      ],
      "properties": {
        "uid": {
          "type": "string"
        }
      },
      "additionalProperties": false
    }
  • grafana_list_alert_rules

    Read summaries of Grafana-managed alert rules; no alert/contact-point changes. This API has no pagination.

    View inputs and schema
    max_results
    Optional
    integer
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "properties": {
        "max_results": {
          "type": "integer",
          "default": 20
        }
      },
      "additionalProperties": false
    }
  • grafana_query_prometheus

    Query Prometheus metrics through an accessible Grafana data source by UID; requires datasource query permission. At most seven days, 1440 points per series and 100 series; does not modify metrics.

    View inputs and schema
    query
    Required
    string
    end_time
    Required
    string
    max_series
    Optional
    integer
    start_time
    Required
    string
    step_seconds
    Optional
    integer
    datasource_uid
    Required
    string
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "required": [
        "datasource_uid",
        "query",
        "start_time",
        "end_time"
      ],
      "properties": {
        "query": {
          "type": "string"
        },
        "end_time": {
          "type": "string"
        },
        "max_series": {
          "type": "integer",
          "default": 20
        },
        "start_time": {
          "type": "string"
        },
        "step_seconds": {
          "type": "integer",
          "default": 300
        },
        "datasource_uid": {
          "type": "string"
        }
      },
      "additionalProperties": false
    }
  • grafana_query_loki

    Read up to 100 Loki log lines over at most seven days through an accessible Grafana data source by UID. Use a LogQL log-stream query, not a metric expression. Log content may be sensitive.

    View inputs and schema
    query
    Required
    string
    end_time
    Required
    string
    start_time
    Required
    string
    max_results
    Optional
    integer
    datasource_uid
    Required
    string
    Full JSON schema, including nested inputs, allowed values, and validation constraints.
    {
      "type": "object",
      "required": [
        "datasource_uid",
        "query",
        "start_time",
        "end_time"
      ],
      "properties": {
        "query": {
          "type": "string"
        },
        "end_time": {
          "type": "string"
        },
        "start_time": {
          "type": "string"
        },
        "max_results": {
          "type": "integer",
          "default": 100
        },
        "datasource_uid": {
          "type": "string"
        }
      },
      "additionalProperties": false
    }
Availability in your MCP client also depends on your account connection, permissions, and active profile.
Connect

Use it from your MCP client

A profile controls which servers and tools your registered clients can use.

  1. Add this server to a profile and set up your personal connection if required.
  2. Select the tools and client instances to allow, then complete the profile’s access review.
  3. Once active, open Connect & test on the profile to get your client configuration and check connectivity.

Register your client in My workspace → My clients and save its token securely. A published listing does not by itself grant runtime access.

Technical details and verification
How a call flows

A round trip: client → registry → server → back

MCP client
Registry gateway
Secured MCP Server
request →← response
Registry guardrails
PolicyProvenanceReflexiveConsentContract
connect
result
MCP client connects to the registry
1 · Authenticate2 · Discover tools3 · Invoke tool
Governance

Gateway controls

Checked controls are enabled for this server’s hosted gateway. Consent and contracts are optional.

○
Policy engine
Only curator-vouched tools are callable — anything else is denied at the gateway.
○
Reflexive core
Per-client behavioral baselines flag anomalous activity.
○
Consent graph
Scoped grants are checked before a tool runs; grants can expire.
○
Contract broker
An active contract sets allowed tools and rate limits per client.
Provenance & verification
Signature
Valid · manifest match
Attested by
Publisher
Hosting
Catalog listing
The registry pins the exact upstream above and records every call made through the gateway in a tamper-evident ledger — visible to the client that made the calls.

User ratings & feedback

Community feedback is separate from security certification. One rating per account; you can update yours.

Loading reviews…

Source & deployment

Not verified

Catalog version: 0.4.0

Submitted source · publisher declaration

Not recorded

Running registry build · operator build record

https://github.com/PureCipher/xsecuremcp2.0

0566b374ddbb720cf486951f791209f7938e5992

Implementation paths and SHA-256 hashes (13)
  • src/purecipher/consumer_aws.py

    9a15442043db8c07a28d9d2037f5ca5ecf79dab0846194e28bd036329d498b10
  • src/purecipher/consumer_bridge.py

    f7247aff802f0869776d5274e5a6c002374235bac8914ffeb6b68ec1536b745e
  • src/purecipher/consumer_bridge_tools.py

    573ad6444316397f78d849f419fcf7a042468b820bf946f2b093c6462b6f5055
  • src/purecipher/consumer_business.py

    b18c519b3a65853180ca07a7234128146fb009b389542659b48dadaebaf3081e
  • src/purecipher/consumer_cloud.py

    08fe162696615e205992db2c4b267c3a65c403cf727314d9e5e941a3597d5ce4
  • src/purecipher/consumer_gmail.py

    3118519c17f7da9f4c9cb84d35bc49ae8d40ab345c897d43421d0e5e7b363680
  • src/purecipher/consumer_google_permissions.py

    99fc2af714c2c177742d3a85b0116e3ed0fe094b2f0ea6e7ded8f9596cfa9c3e
  • src/purecipher/consumer_google_workspace.py

    3799a9e54e74682b187cd57259895fccdb150cfbd7bd57da29fdd32538d806c5
  • src/purecipher/consumer_governance.py

    b1689ec86825ca12eaaf1e95590a97cbec8a46d581debdd0b7772ed3c85b9fea
  • src/purecipher/consumer_oauth.py

    931372166d95e46bcdbe0ef6315741c583c814eff34d4228cccb270336ca312d
  • src/purecipher/consumer_observability.py

    c7d98468ba620e6a2966c1a4fe0ba28ad6feff290396fd2c13f40b7570d5d3e5
  • src/purecipher/consumer_runtime.py

    0ede891d2647e3526dc45876b2b60c064155ce9f42653e748b93003912fb0297
  • src/purecipher/consumer_utilities.py

    e7af046b063271da249c26d06b3d95ab529c77466370f6bbca6550478ee11a19

Comparison covers the registry consumer runtime only. Catalog approval does not deploy code. External upstream deployment identity is not verified; personal connections and credentials are separate.