You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+14-4Lines changed: 14 additions & 4 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -98,20 +98,30 @@ Each launch configuration defined in your workspace becomes a tool named `launch
98
98
99
99
### ⚙️ MCP Options
100
100
101
-
Tasks and launch configurations support an optional `mcp` block for MCP-specific settings. Hover over options in your editor for full documentation.
101
+
Tasks support MCP options inside the `options` block. Launch configurations use them at the top level (you'll see a schema warning, but it works correctly). Hover over options in your editor for full documentation.
Copy file name to clipboardExpand all lines: package.json
+5-1Lines changed: 5 additions & 1 deletion
Original file line number
Diff line number
Diff line change
@@ -64,7 +64,10 @@
64
64
"title": "Ignition MCP",
65
65
"properties": {
66
66
"ignition-mcp.outputLimit": {
67
-
"type": ["integer", "null"],
67
+
"type": [
68
+
"integer",
69
+
"null"
70
+
],
68
71
"default": 20480,
69
72
"minimum": 1024,
70
73
"description": "Default maximum characters to capture from task and debug output. Can be overridden per-task/launch via mcp.outputLimit in tasks.json/launch.json. Set to null for unlimited. Default is 20480 (20KB)."
Copy file name to clipboardExpand all lines: schemas/tasks-mcp.schema.json
+25-20Lines changed: 25 additions & 20 deletions
Original file line number
Diff line number
Diff line change
@@ -7,29 +7,34 @@
7
7
"items": {
8
8
"type": "object",
9
9
"properties": {
10
-
"mcp": {
10
+
"options": {
11
11
"type": "object",
12
-
"description": "MCP-specific options for Ignition MCP extension. Controls how AI assistants interact with this task.",
13
12
"properties": {
14
-
"returnOutput": {
15
-
"type": "string",
16
-
"enum": ["always", "onFailure", "never"],
17
-
"default": "onFailure",
18
-
"markdownDescription": "Controls when terminal output is included in MCP tool responses.\n\n- `always` - Always include full terminal output\n- `onFailure` - Only include output when exit code != 0 (default)\n- `never` - Never include output; use `get_task_output` to retrieve separately\n\nHelps manage AI context size: successful builds don't need hundreds of lines of output, but failed builds should include error details."
19
-
},
20
-
"outputLimit": {
21
-
"type": ["integer", "null"],
22
-
"minimum": 1024,
23
-
"default": 20480,
24
-
"markdownDescription": "Maximum characters to capture from task output. Default is 20480 (20KB). Set to `null` for unlimited. Overrides the global `ignition-mcp.outputLimit` setting."
25
-
},
26
-
"interactive": {
27
-
"type": "boolean",
28
-
"default": false,
29
-
"markdownDescription": "Run task in VS Code's native terminal with full TTY support for interactive input.\n\n**Use when:**\n- Task runs `sudo` commands requiring password input\n- Task prompts for user confirmation (y/n)\n- Commands use `read` or similar interactive input\n- Any command requiring TTY support\n\n**Trade-off:** Output cannot be captured in this mode. The MCP response will show `[Interactive mode: output not captured]`. The terminal will open with focus so the user can interact."
13
+
"mcp": {
14
+
"type": "object",
15
+
"description": "MCP-specific options for Ignition MCP extension. Controls how AI assistants interact with this task.",
16
+
"properties": {
17
+
"returnOutput": {
18
+
"type": "string",
19
+
"enum": ["always", "onFailure", "never"],
20
+
"default": "onFailure",
21
+
"markdownDescription": "Controls when terminal output is included in MCP tool responses.\n\n- `always` - Always include full terminal output\n- `onFailure` - Only include output when exit code != 0 (default)\n- `never` - Never include output; use `get_task_output` to retrieve separately\n\nHelps manage AI context size: successful builds don't need hundreds of lines of output, but failed builds should include error details."
22
+
},
23
+
"outputLimit": {
24
+
"type": ["integer", "null"],
25
+
"minimum": 1024,
26
+
"default": 20480,
27
+
"markdownDescription": "Maximum characters to capture from task output. Default is 20480 (20KB). Set to `null` for unlimited. Overrides the global `ignition-mcp.outputLimit` setting."
28
+
},
29
+
"interactive": {
30
+
"type": "boolean",
31
+
"default": false,
32
+
"markdownDescription": "Run task in VS Code's native terminal with full TTY support for interactive input.\n\n**Use when:**\n- Task runs `sudo` commands requiring password input\n- Task prompts for user confirmation (y/n)\n- Commands use `read` or similar interactive input\n- Any command requiring TTY support\n\n**Trade-off:** Output cannot be captured in this mode. The MCP response will show `[Interactive mode: output not captured]`. The terminal will open with focus so the user can interact."
0 commit comments