@aipermission/mcp 0.2.17 → 0.2.18

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -9,7 +9,7 @@ credentials, or other connector secrets.
9
9
  The gateway is intentionally local-only. Run it on the developer machine and
10
10
  keep the URL on `localhost`; remote systems are connector targets, not places
11
11
  to host the gateway for LAN or internet users. SSH, Postgres, ClickHouse,
12
- Redis / Valkey, RabbitMQ, S3, Docker, and Kubernetes are built-in connectors that use the same
12
+ Redis / Valkey, RabbitMQ, Kafka / Redpanda, S3, Docker, and Kubernetes are built-in connectors that use the same
13
13
  target/profile/action permission model as future connectors.
14
14
 
15
15
  ![AIPermission demo: AI operates through approval-based connector access](https://raw.githubusercontent.com/aipermission/aipermission/main/docs/assets/demo/aipermission-demo.gif)
@@ -66,7 +66,7 @@ The generated MCP config contains a bearer token. Keep it private. For project-l
66
66
  - `cancel_vault_action_request`
67
67
 
68
68
  All integration work goes through connector targets. SSH, Postgres, ClickHouse,
69
- Redis / Valkey, RabbitMQ, S3, Docker, Kubernetes, and future connectors share the same model: target,
69
+ Redis / Valkey, RabbitMQ, Kafka / Redpanda, S3, Docker, Kubernetes, and future connectors share the same model: target,
70
70
  credential profile, connector action, token action permission, approval,
71
71
  history, and audit.
72
72
 
@@ -112,6 +112,12 @@ browser actions such as `overview`, `list_vhosts`, `list_queues`, `get_queue`,
112
112
  previews and published payloads can contain secrets or customer data; use short
113
113
  reasons and prefer approval-required access until the workflow is trusted.
114
114
 
115
+ For Kafka or Redpanda, call `get_connector_actions(target_ref)` to discover
116
+ cluster, topic, consumer-group, lag, and bounded message-read actions. Message
117
+ reads use explicit partition assignment, never join a consumer group, and
118
+ never commit offsets. Payloads can contain sensitive application data, so keep
119
+ these reads in Prompt mode until the workflow is trusted.
120
+
115
121
  For Docker, call `get_connector_actions(target_ref)` to discover bounded
116
122
  actions such as `docker_version`, `list_containers`, `list_images`,
117
123
  `list_networks`, `list_volumes`, `inspect_container`, `container_logs`,
@@ -53,6 +53,10 @@ Peek a RabbitMQ queue after the operator approved payload inspection. Publish
53
53
  RabbitMQ messages only when the operator explicitly asked for a write.
54
54
  ```
55
55
 
56
+ For Kafka / Redpanda, discover topics before describing partitions or reading
57
+ messages. Keep message reads bounded, prefer Prompt because payloads can be
58
+ sensitive, and remember that read actions never join groups or commit offsets.
59
+
56
60
  Avoid vague reasons:
57
61
 
58
62
  ```text
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@aipermission/mcp",
3
- "version": "0.2.17",
3
+ "version": "0.2.18",
4
4
  "mcpName": "io.github.aipermission/aipermission-mcp",
5
5
  "description": "Local-first MCP bridge for the aipermission gateway.",
6
6
  "license": "AGPL-3.0-only",
package/server.json CHANGED
@@ -3,12 +3,12 @@
3
3
  "name": "io.github.aipermission/aipermission-mcp",
4
4
  "title": "AIPermission",
5
5
  "description": "Local-first MCP bridge for the AIPermission gateway.",
6
- "version": "0.2.17",
6
+ "version": "0.2.18",
7
7
  "packages": [
8
8
  {
9
9
  "registryType": "npm",
10
10
  "identifier": "@aipermission/mcp",
11
- "version": "0.2.17",
11
+ "version": "0.2.18",
12
12
  "transport": {
13
13
  "type": "stdio"
14
14
  }