sub-agents-mcp 0.4.0 → 0.4.1
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 +11 -4
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -5,7 +5,7 @@
|
|
|
5
5
|
|
|
6
6
|
Bring Claude Code–style sub-agents to any MCP-compatible tool.
|
|
7
7
|
|
|
8
|
-
This MCP server lets you define task-specific AI agents (like "test-writer" or "code-reviewer") in markdown files, and execute them via Cursor CLI, Claude Code CLI, or
|
|
8
|
+
This MCP server lets you define task-specific AI agents (like "test-writer" or "code-reviewer") in markdown files, and execute them via Cursor CLI, Claude Code, Gemini CLI, or Codex backends.
|
|
9
9
|
|
|
10
10
|
## Why?
|
|
11
11
|
|
|
@@ -14,7 +14,7 @@ Claude Code offers powerful sub-agent workflows—but they're limited to its own
|
|
|
14
14
|
**Concrete benefits:**
|
|
15
15
|
- Define reusable agents once, use them across multiple tools
|
|
16
16
|
- Share agent definitions within teams regardless of IDE choice
|
|
17
|
-
- Leverage Cursor CLI, Claude Code CLI, or
|
|
17
|
+
- Leverage Cursor CLI, Claude Code, Gemini CLI, or Codex capabilities from any MCP client
|
|
18
18
|
|
|
19
19
|
→ [Read the full story](https://dev.to/shinpr/bringing-claude-codes-sub-agents-to-any-mcp-compatible-tool-1hb9)
|
|
20
20
|
|
|
@@ -37,6 +37,7 @@ Claude Code offers powerful sub-agent workflows—but they're limited to its own
|
|
|
37
37
|
- `cursor-agent` CLI (from Cursor)
|
|
38
38
|
- `claude` CLI (from Claude Code)
|
|
39
39
|
- `gemini` CLI (from Gemini CLI)
|
|
40
|
+
- `codex` CLI (from Codex)
|
|
40
41
|
- An MCP-compatible tool (Cursor IDE, Claude Desktop, Windsurf, etc.)
|
|
41
42
|
|
|
42
43
|
## Quick Start
|
|
@@ -90,6 +91,12 @@ gemini
|
|
|
90
91
|
|
|
91
92
|
Note: Gemini CLI uses OAuth authentication. Run `gemini` once to authenticate via browser.
|
|
92
93
|
|
|
94
|
+
**For Codex users:**
|
|
95
|
+
```bash
|
|
96
|
+
# Install Codex
|
|
97
|
+
npm install -g @openai/codex
|
|
98
|
+
```
|
|
99
|
+
|
|
93
100
|
### 3. Configure MCP
|
|
94
101
|
|
|
95
102
|
Add this to your MCP configuration file:
|
|
@@ -391,14 +398,14 @@ The startup overhead is an intentional trade-off: the system favors clarity and
|
|
|
391
398
|
|
|
392
399
|
## How It Works
|
|
393
400
|
|
|
394
|
-
This MCP server acts as a bridge between your AI tool and a supported execution engine (Cursor CLI, Claude Code CLI, or
|
|
401
|
+
This MCP server acts as a bridge between your AI tool and a supported execution engine (Cursor CLI, Claude Code, Gemini CLI, or Codex).
|
|
395
402
|
|
|
396
403
|
**The flow:**
|
|
397
404
|
|
|
398
405
|
1. You configure the MCP server in your client (Cursor, Claude Desktop, etc.)
|
|
399
406
|
2. The client automatically launches `sub-agents-mcp` as a background process when it starts
|
|
400
407
|
3. When your main AI assistant needs a sub-agent, it makes an MCP tool call
|
|
401
|
-
4. The MCP server reads the agent definition (markdown file) and invokes the selected CLI (`cursor-agent`, `claude`, or `
|
|
408
|
+
4. The MCP server reads the agent definition (markdown file) and invokes the selected CLI (`cursor-agent`, `claude`, `gemini`, or `codex`)
|
|
402
409
|
5. The execution engine runs the agent and streams results back through the MCP server
|
|
403
410
|
6. Your main assistant receives the results and continues working
|
|
404
411
|
|
package/package.json
CHANGED