@pi-unipi/subagents 0.2.6 → 0.2.8

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.
Files changed (2) hide show
  1. package/README.md +36 -44
  2. package/package.json +1 -1
package/README.md CHANGED
@@ -1,34 +1,46 @@
1
1
  # @pi-unipi/subagents
2
2
 
3
- Parallel sub-agent execution for [Pi coding agent](https://github.com/badlogic/pi-mono). Spawn background or foreground agents, track activity, and manage concurrent work.
3
+ Parallel execution with file locking. Spawn background or foreground agents to work on tasks concurrently — research files, fix lint errors, run tests while the main agent keeps going.
4
4
 
5
- ## Install
5
+ Two built-in agent types: `explore` for read-only research, `work` for file modifications with transparent locking. Define your own types as markdown files.
6
6
 
7
- ```bash
8
- pi install npm:@pi-unipi/subagents
9
- ```
7
+ ## Commands
10
8
 
11
- Or as part of the full suite:
12
- ```bash
13
- pi install npm:unipi
14
- ```
9
+ Subagents has no user commands. It's an agent tool package — the agent calls it directly.
10
+
11
+ ## Special Triggers
15
12
 
16
- ## Tools
13
+ Workflow skills detect subagents and inject parallel strategies. When `@pi-unipi/subagents` is installed, these skills get enhanced:
14
+
15
+ | Skill | What Changes |
16
+ |-------|--------------|
17
+ | `brainstorm` | Parallel research for different approaches |
18
+ | `document` | Parallel documentation of different modules |
19
+ | `gather-context` | Parallel codebase exploration |
20
+ | `review-work` | Parallel task verification |
21
+ | `scan-issues` | Parallel scanning by category |
22
+ | `work` | Parallel task execution (with file locking) |
23
+
24
+ Subagents registers with the info-screen dashboard, showing active agents and their status. The footer displays agent activity in its extension status segment.
25
+
26
+ ## Agent Tools
17
27
 
18
28
  | Tool | Description |
19
29
  |------|-------------|
20
30
  | `spawn_helper` | Launch a sub-agent for parallel work |
21
31
  | `get_helper_result` | Check status and retrieve results from a background agent |
22
32
 
23
- ## Agent Types
33
+ ### spawn_helper Parameters
24
34
 
25
- | Type | Description |
26
- |------|-------------|
27
- | `explore` | Parallel file reads, research, analysis |
28
- | `work` | Parallel file writes with transparent locking |
29
- | Custom | Define your own in `~/.unipi/config/agents/<name>.md` |
30
-
31
- ## Usage
35
+ | Parameter | Description |
36
+ |-----------|-------------|
37
+ | `type` | Agent type (`explore`, `work`, or custom) |
38
+ | `prompt` | Task for the agent |
39
+ | `description` | Short description (3-5 words) |
40
+ | `run_in_background` | Return immediately, notify on completion |
41
+ | `max_turns` | Max agentic turns before stopping |
42
+ | `model` | Model override (e.g. `"haiku"`, `"sonnet"`) |
43
+ | `thinking` | Thinking level (`off`, `minimal`, `low`, `medium`, `high`, `xhigh`) |
32
44
 
33
45
  ### Foreground (blocks until done)
34
46
 
@@ -57,18 +69,6 @@ spawn_helper(
57
69
  get_helper_result(agent_id: "helper_abc123")
58
70
  ```
59
71
 
60
- ## Options
61
-
62
- | Parameter | Description |
63
- |-----------|-------------|
64
- | `type` | Agent type (`explore`, `work`, or custom) |
65
- | `prompt` | Task for the agent |
66
- | `description` | Short description (3-5 words) |
67
- | `run_in_background` | Return immediately, notify on completion |
68
- | `max_turns` | Max agentic turns before stopping |
69
- | `model` | Model override (e.g. `"haiku"`, `"sonnet"`) |
70
- | `thinking` | Thinking level (`off`, `minimal`, `low`, `medium`, `high`, `xhigh`) |
71
-
72
72
  ## Custom Agent Types
73
73
 
74
74
  Create markdown files defining agent behavior:
@@ -81,7 +81,7 @@ Create markdown files defining agent behavior:
81
81
  <workspace>/.unipi/config/agents/deployer.md
82
82
  ```
83
83
 
84
- ## Configuration
84
+ ## Configurables
85
85
 
86
86
  ```json
87
87
  // ~/.unipi/config/subagents.json
@@ -95,19 +95,11 @@ Create markdown files defining agent behavior:
95
95
  }
96
96
  ```
97
97
 
98
- ## Features
99
-
100
- - **Concurrent execution** run up to N agents simultaneously
101
- - **File locking** transparent locking for parallel writes
102
- - **ESC propagation** kill all agents with ESC
103
- - **Activity tracking** — real-time widget showing agent progress
104
- - **Info screen integration** — agent status in dashboard
105
-
106
- ## Dependencies
107
-
108
- - `@pi-unipi/core` — shared utilities
109
- - `@pi-unipi/workflow` — workflow integration
110
- - `@pi-unipi/info-screen` — dashboard registration
98
+ | Setting | Default | What It Does |
99
+ |---------|---------|--------------|
100
+ | `enabled` | true | Enable/disable subagents |
101
+ | `maxConcurrent` | 3 | Max agents running at once |
102
+ | `types.{name}.enabled` | true | Toggle agent types |
111
103
 
112
104
  ## License
113
105
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@pi-unipi/subagents",
3
- "version": "0.2.6",
3
+ "version": "0.2.8",
4
4
  "description": "Subagents for UniPi — parallel execution, file locking, workflow integration",
5
5
  "type": "module",
6
6
  "main": "dist/index.js",