kiro-agents 1.5.0 → 1.6.1
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 +12 -0
- package/build/npm/dist/agents.md +44 -8
- package/build/npm/dist/aliases.md +4 -5
- package/build/npm/dist/interactions/interaction-styles.md +2 -2
- package/build/npm/dist/protocols/agent-creation.md +282 -0
- package/build/npm/dist/protocols/agent-management.md +173 -0
- package/package.json +2 -2
- package/build/npm/dist/agent-system.md +0 -111
- package/build/npm/dist/modes-system.md +0 -98
- /package/build/npm/dist/protocols/{agent-activation.mdx → agent-activation.md} +0 -0
- /package/build/npm/dist/protocols/{mode-management.mdx → mode-management.md} +0 -0
- /package/build/npm/dist/protocols/{mode-switching.mdx → mode-switching.md} +0 -0
package/README.md
CHANGED
|
@@ -2,6 +2,8 @@
|
|
|
2
2
|
|
|
3
3
|
AI agent system for Kiro IDE. Create specialized agents, switch interaction modes, and reduce cognitive load during development.
|
|
4
4
|
|
|
5
|
+
> For just <kbd>~0.9K</kbd> of **base context** overheat
|
|
6
|
+
|
|
5
7
|
```bash
|
|
6
8
|
npx kiro-agents
|
|
7
9
|
```
|
|
@@ -93,3 +95,13 @@ Traditional AI assistants overwhelm you with context and options. kiro-agents:
|
|
|
93
95
|
## License
|
|
94
96
|
|
|
95
97
|
MIT
|
|
98
|
+
|
|
99
|
+
## Contributing
|
|
100
|
+
|
|
101
|
+
We welcome contributions! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.
|
|
102
|
+
|
|
103
|
+
**Important**: Do not modify files in `power/` directory. This is auto-generated during release. See contributing guide for details.
|
|
104
|
+
|
|
105
|
+
## License
|
|
106
|
+
|
|
107
|
+
MIT
|
package/build/npm/dist/agents.md
CHANGED
|
@@ -6,6 +6,26 @@ keywords: ["agents", "management", "create", "activate", "interactive"]
|
|
|
6
6
|
|
|
7
7
|
# Agent Management
|
|
8
8
|
|
|
9
|
+
## Parameter Detection
|
|
10
|
+
|
|
11
|
+
**CRITICAL: Check if user provided agent name as parameter**
|
|
12
|
+
|
|
13
|
+
If the user's message contains `/agents {agent_name}` with a specific agent name (e.g., `/agents kiro-master`):
|
|
14
|
+
|
|
15
|
+
1. **DO NOT execute the interactive management flow below**
|
|
16
|
+
2. **INSTEAD, execute the instruction alias for agent activation:**
|
|
17
|
+
- Read `.kiro/agents/{agent_name}.md` into context
|
|
18
|
+
- Read #[[file:protocols/agent-activation.md]] into context
|
|
19
|
+
- Follow all steps from the "Agent Activation Steps" section in agent-activation.md
|
|
20
|
+
- Use `{agent_name}` as the agent identifier throughout the protocol
|
|
21
|
+
3. **Stop processing this document** - The agent activation protocol takes over
|
|
22
|
+
|
|
23
|
+
**Only continue with interactive management flow below if user typed `/agents` without parameters.**
|
|
24
|
+
|
|
25
|
+
---
|
|
26
|
+
|
|
27
|
+
## Interactive Management Mode
|
|
28
|
+
|
|
9
29
|
You are now in **agent management mode** using chit-chat interaction protocol.
|
|
10
30
|
|
|
11
31
|
## Agent Management Steps
|
|
@@ -37,9 +57,9 @@ Begin with a diff block showing:
|
|
|
37
57
|
Execute `listDirectory` on `.kiro/agents/`:
|
|
38
58
|
|
|
39
59
|
**If directory doesn't exist OR directory is empty:**
|
|
40
|
-
1. **Read
|
|
60
|
+
1. **Read #[[file:protocols/agent-creation.md]] into context** - Load agent creation protocol
|
|
41
61
|
2. Create `.kiro/agents/kiro-master.md` agent automatically using the description from "Initial Agent" section
|
|
42
|
-
3. Follow the agent definition structure from agent-creation.
|
|
62
|
+
3. Follow the agent definition structure from agent-creation.md protocol
|
|
43
63
|
4. Show diff block indicating setup completion
|
|
44
64
|
5. Continue to Step 3 with kiro-master as available agent
|
|
45
65
|
|
|
@@ -89,8 +109,8 @@ Based on user selection:
|
|
|
89
109
|
|
|
90
110
|
#### Option 2 - Create New Agent
|
|
91
111
|
|
|
92
|
-
- **Read
|
|
93
|
-
- Follow all steps from the "Agent Creation Steps" section in agent-creation.
|
|
112
|
+
- **Read #[[file:protocols/agent-creation.md]] into context** - Load agent creation protocol
|
|
113
|
+
- Follow all steps from the "Agent Creation Steps" section in agent-creation.md
|
|
94
114
|
- Start agent creation workflow with interactive wizard
|
|
95
115
|
- Ask for agent type (code-focused, documentation, testing, etc.)
|
|
96
116
|
- Collect agent name, description, capabilities
|
|
@@ -100,7 +120,7 @@ Based on user selection:
|
|
|
100
120
|
|
|
101
121
|
#### Option 3 - Manage Existing Agent
|
|
102
122
|
|
|
103
|
-
- **Read
|
|
123
|
+
- **Read #[[file:protocols/agent-creation.md]] into context** - Load agent structure reference
|
|
104
124
|
- Show numbered list of available agents
|
|
105
125
|
- User selects agent to manage
|
|
106
126
|
- Offer management options:
|
|
@@ -133,7 +153,11 @@ AGENT COMMANDS
|
|
|
133
153
|
/agents {name} Activate specific agent
|
|
134
154
|
/agents Interactive agent management
|
|
135
155
|
|
|
136
|
-
|
|
156
|
+
MODE COMMANDS (see modes-system.md)
|
|
157
|
+
/modes {name} Switch Kiro mode (vibe/spec)
|
|
158
|
+
/modes Interactive mode management
|
|
159
|
+
/strict {state} Control strict mode (on/off)
|
|
160
|
+
/strict Interactive strict mode control
|
|
137
161
|
|
|
138
162
|
Note: Commands are defined in steering documents.
|
|
139
163
|
New commands can be added via Instruction Alias pattern.
|
|
@@ -151,7 +175,7 @@ Explain agent system:
|
|
|
151
175
|
#### Option 7 - Exit
|
|
152
176
|
|
|
153
177
|
- Confirm exit from agent management mode
|
|
154
|
-
- Return to normal
|
|
178
|
+
- Return to normal interaction
|
|
155
179
|
- Preserve any changes made
|
|
156
180
|
|
|
157
181
|
### Step 5: Maintain Chit-Chat Mode
|
|
@@ -185,4 +209,16 @@ This description is used when creating `.kiro/agents/kiro-master.md` during auto
|
|
|
185
209
|
|
|
186
210
|
---
|
|
187
211
|
|
|
188
|
-
**
|
|
212
|
+
**Quick Commands:**
|
|
213
|
+
|
|
214
|
+
```
|
|
215
|
+
/agents Interactive agent management
|
|
216
|
+
|
|
217
|
+
MODE COMMANDS (see modes-system.md)
|
|
218
|
+
/modes {name} Switch Kiro mode (vibe/spec)
|
|
219
|
+
/modes Interactive mode management
|
|
220
|
+
/strict {state} Control strict mode (on/off)
|
|
221
|
+
/strict Interactive strict mode control
|
|
222
|
+
```
|
|
223
|
+
|
|
224
|
+
Note: Commands are defined in steering documents.
|
|
@@ -89,15 +89,14 @@ You are now activating the **{agent_name}** agent.
|
|
|
89
89
|
|
|
90
90
|
**Load and execute activation protocol:**
|
|
91
91
|
1. Read `.kiro/agents/{agent_name}.md` into context
|
|
92
|
-
2. Read
|
|
93
|
-
3. Follow all steps from the "Agent Activation Steps" section in agent-activation.
|
|
92
|
+
2. Read #[[file:protocols/agent-activation.md]] into context
|
|
93
|
+
3. Follow all steps from the "Agent Activation Steps" section in agent-activation.md
|
|
94
94
|
4. Use `{agent_name}` as the agent identifier throughout the protocol
|
|
95
95
|
</definition>
|
|
96
96
|
</alias>
|
|
97
97
|
|
|
98
98
|
This alias enables users to activate any agent with `/agents {name}` syntax.
|
|
99
99
|
|
|
100
|
-
|
|
101
100
|
## Mode System Alias
|
|
102
101
|
|
|
103
102
|
The mode switching command uses parameter substitution to load mode definitions dynamically:
|
|
@@ -111,8 +110,8 @@ You are now switching to **{mode_name} mode**.
|
|
|
111
110
|
|
|
112
111
|
**Load and execute mode switching protocol:**
|
|
113
112
|
1. Read `kiro-{mode_name}-mode.md` from agent-system directory into context
|
|
114
|
-
2. Read
|
|
115
|
-
3. Follow all steps from the "Mode Switch Steps" section in mode-switching.
|
|
113
|
+
2. Read #[[file:protocols/mode-switching.md]] into context
|
|
114
|
+
3. Follow all steps from the "Mode Switch Steps" section in mode-switching.md
|
|
116
115
|
4. Use `{mode_name}` as the mode identifier throughout the protocol
|
|
117
116
|
</definition>
|
|
118
117
|
</alias>
|
|
@@ -16,7 +16,7 @@ inclusion: manual
|
|
|
16
16
|
- Visual formatting (bold, code blocks)
|
|
17
17
|
- Context recovery system
|
|
18
18
|
|
|
19
|
-
**Example agents**:
|
|
19
|
+
**Example agents**: project-manager
|
|
20
20
|
|
|
21
21
|
**When to use**:
|
|
22
22
|
|
|
@@ -103,7 +103,7 @@ inclusion: manual
|
|
|
103
103
|
- Simple tasks → Direct execution
|
|
104
104
|
- Complex tasks → Chit-chat or consultative
|
|
105
105
|
|
|
106
|
-
**Example agents**:
|
|
106
|
+
**Example agents**: project-manager-hybrid
|
|
107
107
|
|
|
108
108
|
**When to use**:
|
|
109
109
|
|
|
@@ -0,0 +1,282 @@
|
|
|
1
|
+
# Agent Creation Protocol
|
|
2
|
+
|
|
3
|
+
This file contains the detailed instructions for creating new agents. It is referenced during agent creation workflow in agent management mode.
|
|
4
|
+
|
|
5
|
+
## Agent Creation Steps
|
|
6
|
+
|
|
7
|
+
When creating a new agent:
|
|
8
|
+
|
|
9
|
+
### Step 1: Initialize Creation Wizard
|
|
10
|
+
|
|
11
|
+
Use chit-chat mode for the creation process:
|
|
12
|
+
|
|
13
|
+
```diff
|
|
14
|
+
👉 Agent creation wizard
|
|
15
|
+
⏳ Gathering agent information
|
|
16
|
+
```
|
|
17
|
+
|
|
18
|
+
**Current Focus**: Creating new agent
|
|
19
|
+
|
|
20
|
+
### Step 2: Collect Agent Information
|
|
21
|
+
|
|
22
|
+
Guide user through agent definition with numbered choices:
|
|
23
|
+
|
|
24
|
+
#### 2.1: Agent Type Selection
|
|
25
|
+
|
|
26
|
+
**What type of agent would you like to create?**
|
|
27
|
+
|
|
28
|
+
1. **Specialist** - Focused on specific domain (e.g., refactoring, testing)
|
|
29
|
+
- Deep expertise in narrow area
|
|
30
|
+
- Clear boundaries of responsibility
|
|
31
|
+
- Best for: Specialized tasks requiring domain expertise
|
|
32
|
+
|
|
33
|
+
2. **Generalist** - Broad capabilities across domains
|
|
34
|
+
- Flexible and adaptable
|
|
35
|
+
- Good for exploration and discovery
|
|
36
|
+
- Best for: General-purpose assistance
|
|
37
|
+
|
|
38
|
+
3. **Coordinator** - Manages other agents
|
|
39
|
+
- Orchestrates complex workflows
|
|
40
|
+
- Handles agent handoffs
|
|
41
|
+
- Best for: Multi-agent workflows
|
|
42
|
+
|
|
43
|
+
#### 2.2: Agent Name
|
|
44
|
+
|
|
45
|
+
**What should we name this agent?**
|
|
46
|
+
|
|
47
|
+
Requirements:
|
|
48
|
+
- Lowercase with hyphens (e.g., `refactor-architect`)
|
|
49
|
+
- No spaces or special characters
|
|
50
|
+
- Descriptive of agent's purpose
|
|
51
|
+
- Must be unique (check `.kiro/agents/` directory)
|
|
52
|
+
|
|
53
|
+
#### 2.3: Agent Description
|
|
54
|
+
|
|
55
|
+
**Provide a brief description (1-2 sentences):**
|
|
56
|
+
|
|
57
|
+
This will be shown in agent listings and help users understand when to use this agent.
|
|
58
|
+
|
|
59
|
+
Example: "Specialized in code refactoring with focus on design patterns, SOLID principles, and maintainability improvements."
|
|
60
|
+
|
|
61
|
+
#### 2.4: Core Responsibilities
|
|
62
|
+
|
|
63
|
+
**What are the agent's primary responsibilities? (3-5 items)**
|
|
64
|
+
|
|
65
|
+
Example:
|
|
66
|
+
- Analyze code structure and identify refactoring opportunities
|
|
67
|
+
- Apply design patterns appropriately
|
|
68
|
+
- Improve code maintainability and readability
|
|
69
|
+
- Ensure SOLID principles compliance
|
|
70
|
+
|
|
71
|
+
#### 2.5: Capabilities
|
|
72
|
+
|
|
73
|
+
**What specific capabilities should this agent have?**
|
|
74
|
+
|
|
75
|
+
Group by category. Example:
|
|
76
|
+
|
|
77
|
+
**Code Analysis:**
|
|
78
|
+
- Detect code smells and anti-patterns
|
|
79
|
+
- Identify coupling and cohesion issues
|
|
80
|
+
- Analyze complexity metrics
|
|
81
|
+
|
|
82
|
+
**Refactoring:**
|
|
83
|
+
- Extract methods and classes
|
|
84
|
+
- Apply design patterns
|
|
85
|
+
- Simplify conditional logic
|
|
86
|
+
|
|
87
|
+
#### 2.6: Interaction Protocol
|
|
88
|
+
|
|
89
|
+
**How should this agent interact with users?**
|
|
90
|
+
|
|
91
|
+
1. **Formal** - Professional, structured responses
|
|
92
|
+
2. **Casual** - Conversational, friendly tone
|
|
93
|
+
3. **Technical** - Precise, detailed explanations
|
|
94
|
+
4. **Chit-chat mode** - Interactive with diff blocks and numbered choices
|
|
95
|
+
|
|
96
|
+
**Confirmation requirements:**
|
|
97
|
+
- Always confirm before changes
|
|
98
|
+
- Never confirm (trusted operations)
|
|
99
|
+
- Confirm only risky operations
|
|
100
|
+
|
|
101
|
+
#### 2.7: Mandatory Protocols
|
|
102
|
+
|
|
103
|
+
**What rules MUST this agent follow?**
|
|
104
|
+
|
|
105
|
+
Example:
|
|
106
|
+
1. **Always analyze before refactoring** - Never refactor without understanding impact
|
|
107
|
+
2. **Preserve functionality** - Ensure tests pass after refactoring
|
|
108
|
+
3. **Document changes** - Explain why refactoring was done
|
|
109
|
+
|
|
110
|
+
#### 2.8: Workflows
|
|
111
|
+
|
|
112
|
+
**Define step-by-step processes for common tasks:**
|
|
113
|
+
|
|
114
|
+
Example workflow:
|
|
115
|
+
|
|
116
|
+
**Refactor Code:**
|
|
117
|
+
1. Analyze current code structure
|
|
118
|
+
2. Identify refactoring opportunities
|
|
119
|
+
3. Propose refactoring plan
|
|
120
|
+
4. Get user confirmation
|
|
121
|
+
5. Apply refactoring
|
|
122
|
+
6. Verify tests pass
|
|
123
|
+
7. Document changes
|
|
124
|
+
|
|
125
|
+
### Step 3: Generate Agent Definition File
|
|
126
|
+
|
|
127
|
+
Create `.kiro/agents/{agent-name}.md` with this structure:
|
|
128
|
+
|
|
129
|
+
```markdown
|
|
130
|
+
---
|
|
131
|
+
name: {agent-name}
|
|
132
|
+
type: {specialist|generalist|coordinator}
|
|
133
|
+
description: {brief description}
|
|
134
|
+
version: 1.0.0
|
|
135
|
+
---
|
|
136
|
+
|
|
137
|
+
# {Agent Name}
|
|
138
|
+
|
|
139
|
+
{Brief overview of agent purpose and capabilities}
|
|
140
|
+
|
|
141
|
+
## Core Responsibilities
|
|
142
|
+
|
|
143
|
+
{List of primary responsibilities}
|
|
144
|
+
|
|
145
|
+
## Capabilities
|
|
146
|
+
|
|
147
|
+
{Detailed list organized by category}
|
|
148
|
+
|
|
149
|
+
## Interaction Protocol
|
|
150
|
+
|
|
151
|
+
{How agent responds to user input}
|
|
152
|
+
|
|
153
|
+
**Response Style:** {formal|casual|technical|chit-chat}
|
|
154
|
+
|
|
155
|
+
**Formatting Preferences:**
|
|
156
|
+
- {Code blocks, lists, tables, etc.}
|
|
157
|
+
|
|
158
|
+
**Confirmation Requirements:**
|
|
159
|
+
- {Always, never, when risky}
|
|
160
|
+
|
|
161
|
+
**Error Handling:**
|
|
162
|
+
- {How agent handles errors}
|
|
163
|
+
|
|
164
|
+
## Mandatory Protocols
|
|
165
|
+
|
|
166
|
+
{Rules agent MUST follow with rationale}
|
|
167
|
+
|
|
168
|
+
## Workflows
|
|
169
|
+
|
|
170
|
+
{Step-by-step processes for common tasks}
|
|
171
|
+
|
|
172
|
+
## Examples
|
|
173
|
+
|
|
174
|
+
### Example 1: {Scenario Name}
|
|
175
|
+
|
|
176
|
+
**User:** {Example user input}
|
|
177
|
+
**Agent:** {Example agent response}
|
|
178
|
+
|
|
179
|
+
### Example 2: {Scenario Name}
|
|
180
|
+
|
|
181
|
+
**User:** {Example user input}
|
|
182
|
+
**Agent:** {Example agent response}
|
|
183
|
+
|
|
184
|
+
## Integration Points
|
|
185
|
+
|
|
186
|
+
### Required Steering Documents
|
|
187
|
+
- `{steering-name}.md` - {Why it's required}
|
|
188
|
+
|
|
189
|
+
### Optional Steering Documents
|
|
190
|
+
- `{steering-name}.md` - {When to load}
|
|
191
|
+
|
|
192
|
+
## Conflict Priorities
|
|
193
|
+
|
|
194
|
+
When conflicts arise, this agent prioritizes:
|
|
195
|
+
|
|
196
|
+
1. User safety and data integrity
|
|
197
|
+
2. Agent mandatory protocols
|
|
198
|
+
3. Project-specific requirements
|
|
199
|
+
4. Performance and efficiency
|
|
200
|
+
5. Code style and conventions
|
|
201
|
+
|
|
202
|
+
## Best Practices
|
|
203
|
+
|
|
204
|
+
- {Best practice 1 for using this agent}
|
|
205
|
+
- {Best practice 2 for using this agent}
|
|
206
|
+
- {Best practice 3 for using this agent}
|
|
207
|
+
|
|
208
|
+
## Advanced Features
|
|
209
|
+
|
|
210
|
+
- {Advanced feature 1 description}
|
|
211
|
+
- {Advanced feature 2 description}
|
|
212
|
+
|
|
213
|
+
## Error Handling
|
|
214
|
+
|
|
215
|
+
- **Syntax errors** - {Approach description}
|
|
216
|
+
- **Logic errors** - {Approach description}
|
|
217
|
+
- **Ambiguous input** - {Approach description}
|
|
218
|
+
|
|
219
|
+
## Success Metrics
|
|
220
|
+
|
|
221
|
+
- {Metric 1 description}
|
|
222
|
+
- {Metric 2 description}
|
|
223
|
+
```
|
|
224
|
+
|
|
225
|
+
### Step 4: Validate Agent Definition
|
|
226
|
+
|
|
227
|
+
Check that the generated agent file:
|
|
228
|
+
|
|
229
|
+
1. **Has valid YAML frontmatter** with required fields
|
|
230
|
+
2. **Contains all required sections**:
|
|
231
|
+
- Core Responsibilities
|
|
232
|
+
- Capabilities
|
|
233
|
+
- Interaction Protocol
|
|
234
|
+
- Mandatory Protocols
|
|
235
|
+
- Workflows
|
|
236
|
+
- Examples
|
|
237
|
+
3. **Has no syntax errors** in markdown
|
|
238
|
+
4. **Follows naming conventions** (lowercase with hyphens)
|
|
239
|
+
5. **Is unique** (no existing agent with same name)
|
|
240
|
+
|
|
241
|
+
### Step 5: Confirm and Finalize
|
|
242
|
+
|
|
243
|
+
Show summary to user:
|
|
244
|
+
|
|
245
|
+
```diff
|
|
246
|
+
👉 Agent creation complete
|
|
247
|
+
✅ Agent definition created
|
|
248
|
+
```
|
|
249
|
+
|
|
250
|
+
**Agent Created:** `{agent-name}`
|
|
251
|
+
|
|
252
|
+
**Location:** `.kiro/agents/{agent-name}.md`
|
|
253
|
+
|
|
254
|
+
**Summary:**
|
|
255
|
+
- Type: {type}
|
|
256
|
+
- Responsibilities: {count} defined
|
|
257
|
+
- Capabilities: {count} defined
|
|
258
|
+
- Workflows: {count} defined
|
|
259
|
+
|
|
260
|
+
**What would you like to do next?**
|
|
261
|
+
|
|
262
|
+
1. **Activate this agent** - Start using it immediately
|
|
263
|
+
2. **View agent definition** - Review the generated file
|
|
264
|
+
3. **Edit agent** - Make modifications
|
|
265
|
+
4. **Create another agent** - Start new creation wizard
|
|
266
|
+
5. **Return to agent management** - Back to main menu
|
|
267
|
+
6. **Exit** - Return to normal mode
|
|
268
|
+
|
|
269
|
+
### Step 6: Post-Creation Actions
|
|
270
|
+
|
|
271
|
+
Based on user choice:
|
|
272
|
+
|
|
273
|
+
- **Activate**: Execute `/agents {agent-name}` command
|
|
274
|
+
- **View**: Display full agent definition with formatting
|
|
275
|
+
- **Edit**: Open agent file for modifications
|
|
276
|
+
- **Create another**: Restart creation wizard
|
|
277
|
+
- **Return**: Go back to agent management menu
|
|
278
|
+
- **Exit**: Return to normal interaction
|
|
279
|
+
|
|
280
|
+
---
|
|
281
|
+
|
|
282
|
+
**Agent creation wizard ready. Begin collecting information.**
|
|
@@ -0,0 +1,173 @@
|
|
|
1
|
+
# Agent Management Protocol
|
|
2
|
+
|
|
3
|
+
This file contains the detailed instructions for interactive agent management. It is referenced when `/agents` is executed without parameters.
|
|
4
|
+
|
|
5
|
+
## Agent Management Steps
|
|
6
|
+
|
|
7
|
+
When entering agent management mode:
|
|
8
|
+
|
|
9
|
+
### Step 1: Activate Chit-Chat Mode
|
|
10
|
+
|
|
11
|
+
Load and apply protocols from `chit-chat.md` steering document:
|
|
12
|
+
|
|
13
|
+
- **Use diff blocks** to show progress and current state
|
|
14
|
+
- **Provide numbered choice lists** (4-8 options, up to 16 if needed)
|
|
15
|
+
- **Maintain single focus** per message
|
|
16
|
+
- **Use visual formatting** (bold, code blocks, lists)
|
|
17
|
+
- **Optimize for ADHD-C** (minimal cognitive load)
|
|
18
|
+
|
|
19
|
+
Begin with a diff block showing:
|
|
20
|
+
```diff
|
|
21
|
+
👉 Agent management mode
|
|
22
|
+
⏳ [Current operation]
|
|
23
|
+
```
|
|
24
|
+
|
|
25
|
+
### Step 2: Scan Agents Directory
|
|
26
|
+
|
|
27
|
+
**CRITICAL: You MUST execute `listDirectory` tool on `.kiro/agents/` to get the actual list of agents.**
|
|
28
|
+
|
|
29
|
+
**DO NOT rely on open editor files or context - always scan the directory explicitly.**
|
|
30
|
+
|
|
31
|
+
Execute `listDirectory` on `.kiro/agents/`:
|
|
32
|
+
|
|
33
|
+
**If directory doesn't exist OR directory is empty:**
|
|
34
|
+
1. **Read #[[file:protocols/agent-creation.md]] into context** - Load agent creation protocol
|
|
35
|
+
2. Create `.kiro/agents/kiro-master.md` agent automatically using the description from "Initial Agent" section
|
|
36
|
+
3. Follow the agent definition structure from agent-creation.md protocol
|
|
37
|
+
4. Show diff block indicating setup completion
|
|
38
|
+
5. Continue to Step 3 with kiro-master as available agent
|
|
39
|
+
|
|
40
|
+
**If directory exists with agents:**
|
|
41
|
+
1. **Execute `listDirectory` tool** to get complete list of `.md` files
|
|
42
|
+
2. Extract agent names from filenames (remove `.md` extension)
|
|
43
|
+
3. Read frontmatter metadata for descriptions
|
|
44
|
+
4. Categorize by agent type if metadata available
|
|
45
|
+
5. **List ALL agents found** - do not filter or assume based on context
|
|
46
|
+
|
|
47
|
+
### Step 3: Present Agent Selection
|
|
48
|
+
|
|
49
|
+
Use this response structure:
|
|
50
|
+
|
|
51
|
+
```diff
|
|
52
|
+
👉 Agent management mode
|
|
53
|
+
⏳ Agent selection or management
|
|
54
|
+
```
|
|
55
|
+
|
|
56
|
+
**Current Focus**: Agent selection and management
|
|
57
|
+
|
|
58
|
+
**Available agents in `.kiro/agents/`:**
|
|
59
|
+
|
|
60
|
+
[List each agent with format:]
|
|
61
|
+
- **{agent-name}** - {description from frontmatter or .md}
|
|
62
|
+
|
|
63
|
+
**What would you like to do?**
|
|
64
|
+
|
|
65
|
+
1. **Activate agent** - Choose agent to activate immediately
|
|
66
|
+
2. **Create new agent** - Start agent creation wizard
|
|
67
|
+
3. **Manage existing agent** - Modify agent configuration
|
|
68
|
+
4. **View agent details** - Show full agent definition
|
|
69
|
+
5. **List all commands** - Show available slash commands
|
|
70
|
+
6. **Agent system help** - Explain how agents work
|
|
71
|
+
7. **Exit** - Return to normal mode
|
|
72
|
+
|
|
73
|
+
### Step 4: Handle User Choice
|
|
74
|
+
|
|
75
|
+
Based on user selection:
|
|
76
|
+
|
|
77
|
+
#### Option 1 - Activate Agent
|
|
78
|
+
|
|
79
|
+
- Show numbered list of available agents
|
|
80
|
+
- User selects agent by number
|
|
81
|
+
- Execute `/agents {selected_agent_name}` command
|
|
82
|
+
- Agent activates immediately
|
|
83
|
+
|
|
84
|
+
#### Option 2 - Create New Agent
|
|
85
|
+
|
|
86
|
+
- **Read #[[file:protocols/agent-creation.md]] into context** - Load agent creation protocol
|
|
87
|
+
- Follow all steps from the "Agent Creation Steps" section in agent-creation.md
|
|
88
|
+
- Start agent creation workflow with interactive wizard
|
|
89
|
+
- Ask for agent type (code-focused, documentation, testing, etc.)
|
|
90
|
+
- Collect agent name, description, capabilities
|
|
91
|
+
- Generate `.md` file with all agent definition following protocol structure
|
|
92
|
+
- Validate agent definition per protocol
|
|
93
|
+
- Offer to activate new agent
|
|
94
|
+
|
|
95
|
+
#### Option 3 - Manage Existing Agent
|
|
96
|
+
|
|
97
|
+
- **Read #[[file:protocols/agent-creation.md]] into context** - Load agent structure reference
|
|
98
|
+
- Show numbered list of available agents
|
|
99
|
+
- User selects agent to manage
|
|
100
|
+
- Offer management options:
|
|
101
|
+
- Modify capabilities
|
|
102
|
+
- Update instructions
|
|
103
|
+
- Change interaction protocol
|
|
104
|
+
- Add/remove integrations
|
|
105
|
+
- Delete agent
|
|
106
|
+
- Use agent definition structure from protocol as reference for modifications
|
|
107
|
+
- Apply changes to agent `.md` file
|
|
108
|
+
- Validate changes against protocol requirements
|
|
109
|
+
- Reload agent if currently active
|
|
110
|
+
|
|
111
|
+
#### Option 4 - View Agent Details
|
|
112
|
+
|
|
113
|
+
- Show numbered list of available agents
|
|
114
|
+
- User selects agent to view
|
|
115
|
+
- Display full `.md` content with formatting
|
|
116
|
+
- Show agent protocols and capabilities summary
|
|
117
|
+
- Offer to activate or manage agent
|
|
118
|
+
|
|
119
|
+
#### Option 5 - List All Commands
|
|
120
|
+
|
|
121
|
+
Show all available slash commands in the system:
|
|
122
|
+
|
|
123
|
+
```
|
|
124
|
+
Available Commands:
|
|
125
|
+
|
|
126
|
+
AGENT COMMANDS
|
|
127
|
+
/agents {name} Activate specific agent
|
|
128
|
+
/agents Interactive agent management
|
|
129
|
+
|
|
130
|
+
MODE COMMANDS (see modes-system.md)
|
|
131
|
+
/modes {name} Switch Kiro mode (vibe/spec)
|
|
132
|
+
/modes Interactive mode management
|
|
133
|
+
/strict {state} Control strict mode (on/off)
|
|
134
|
+
/strict Interactive strict mode control
|
|
135
|
+
|
|
136
|
+
Note: Commands are defined in steering documents.
|
|
137
|
+
New commands can be added via Instruction Alias pattern.
|
|
138
|
+
```
|
|
139
|
+
|
|
140
|
+
#### Option 6 - Agent System Help
|
|
141
|
+
|
|
142
|
+
Explain agent system:
|
|
143
|
+
- Agent system architecture
|
|
144
|
+
- Command usage examples
|
|
145
|
+
- Agent file structure
|
|
146
|
+
- How to create custom agents
|
|
147
|
+
- Best practices for agent usage
|
|
148
|
+
|
|
149
|
+
#### Option 7 - Exit
|
|
150
|
+
|
|
151
|
+
- Confirm exit from agent management mode
|
|
152
|
+
- Return to normal interaction
|
|
153
|
+
- Preserve any changes made
|
|
154
|
+
|
|
155
|
+
### Step 5: Maintain Chit-Chat Mode
|
|
156
|
+
|
|
157
|
+
Continue using diff blocks and numbered choices throughout agent management session.
|
|
158
|
+
|
|
159
|
+
After each action:
|
|
160
|
+
- Update diff block with progress
|
|
161
|
+
- Show current focus
|
|
162
|
+
- Provide next action choices
|
|
163
|
+
- Allow going back or canceling
|
|
164
|
+
|
|
165
|
+
**Context Preservation:**
|
|
166
|
+
- Use diff blocks to show progress
|
|
167
|
+
- Preserve user decisions
|
|
168
|
+
- Allow going back to previous step
|
|
169
|
+
- Enable canceling operations
|
|
170
|
+
|
|
171
|
+
---
|
|
172
|
+
|
|
173
|
+
**Agent management mode active. Present choices to user.**
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "kiro-agents",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.6.1",
|
|
4
4
|
"description": "Advanced AI agent system for Kiro IDE - Create specialized AI agents, switch interaction modes, and enhance development workflows with minimal cognitive overhead",
|
|
5
5
|
"homepage": "https://github.com/Theadd/kiro-agents#readme",
|
|
6
6
|
"repository": {
|
|
@@ -52,7 +52,7 @@
|
|
|
52
52
|
"author": "R. Beltran",
|
|
53
53
|
"license": "MIT",
|
|
54
54
|
"devDependencies": {
|
|
55
|
-
"@changesets/cli": "
|
|
55
|
+
"@changesets/cli": "2.27.9",
|
|
56
56
|
"@types/bun": "latest"
|
|
57
57
|
},
|
|
58
58
|
"peerDependencies": {
|
|
@@ -1,111 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
inclusion: always
|
|
3
|
-
description: Generic agent activation system with instruction commands for seamless agent management
|
|
4
|
-
---
|
|
5
|
-
|
|
6
|
-
# Agent System
|
|
7
|
-
|
|
8
|
-
Generic agent activation and management system for Kiro using instruction commands with parameter support.
|
|
9
|
-
|
|
10
|
-
**For complete documentation, examples, and troubleshooting, see `docs/agent-system-guide.md`**
|
|
11
|
-
|
|
12
|
-
## Agent Commands
|
|
13
|
-
|
|
14
|
-
**Note:** This system uses the Instruction Alias pattern defined in `aliases.md`.
|
|
15
|
-
|
|
16
|
-
### Command 1: Activate Specific Agent
|
|
17
|
-
|
|
18
|
-
<alias>
|
|
19
|
-
<trigger>/agents {agent_name}</trigger>
|
|
20
|
-
<definition>
|
|
21
|
-
## Agent Activation: {agent_name}
|
|
22
|
-
|
|
23
|
-
You are now activating the **{agent_name}** agent.
|
|
24
|
-
|
|
25
|
-
**Load and execute activation protocol:**
|
|
26
|
-
1. Read `.kiro/agents/{agent_name}.md` into context
|
|
27
|
-
2. Read `~/.kiro/steering/kiro-agents/protocols/agent-activation.mdx` into context
|
|
28
|
-
3. Follow all steps from the "Agent Activation Steps" section in agent-activation.mdx
|
|
29
|
-
4. Use `{agent_name}` as the agent identifier throughout the protocol
|
|
30
|
-
</definition>
|
|
31
|
-
</alias>
|
|
32
|
-
|
|
33
|
-
### Command 2: Interactive Agent Management
|
|
34
|
-
|
|
35
|
-
For interactive agent management, use the `/agents` slash command (without parameters).
|
|
36
|
-
This provides a visual interface for:
|
|
37
|
-
- Viewing all available agents
|
|
38
|
-
- Creating new agents
|
|
39
|
-
- Managing existing agents
|
|
40
|
-
- Viewing agent details
|
|
41
|
-
|
|
42
|
-
**Usage:**
|
|
43
|
-
- Interactive: `/agents` (no parameters, loads `agents.md` steering file)
|
|
44
|
-
- Direct: `/agents {name}` (with agent name, uses alias above)
|
|
45
|
-
|
|
46
|
-
## Quick Reference
|
|
47
|
-
|
|
48
|
-
### Activation Protocol Summary
|
|
49
|
-
|
|
50
|
-
When `/agents {name}` is executed:
|
|
51
|
-
|
|
52
|
-
1. **Discover** - Check `.kiro/agents/{name}.md` exists
|
|
53
|
-
2. **Load** - Read agent definition and `strict-mode.md`
|
|
54
|
-
3. **Apply** - Follow mandatory protocols, load required steering
|
|
55
|
-
4. **Assume** - Become agent completely, maintain role
|
|
56
|
-
5. **Begin** - Start interaction per agent's protocol
|
|
57
|
-
|
|
58
|
-
### Management Protocol Summary
|
|
59
|
-
|
|
60
|
-
When `/agents` is executed:
|
|
61
|
-
|
|
62
|
-
1. **Activate chit-chat** - Load chit-chat.md, use diff blocks
|
|
63
|
-
2. **Scan directory** - List `.md` files in `.kiro/agents/`
|
|
64
|
-
3. **Present choices** - Show agents, offer operations
|
|
65
|
-
4. **Handle operations** - Activation, creation, management, viewing
|
|
66
|
-
5. **Maintain context** - Track progress, preserve decisions
|
|
67
|
-
|
|
68
|
-
## Agent Definition Requirements
|
|
69
|
-
|
|
70
|
-
**File location:** `.kiro/agents/{agent-name}.md`
|
|
71
|
-
|
|
72
|
-
**Required sections:**
|
|
73
|
-
- Frontmatter (name, type, description, version)
|
|
74
|
-
- Core Responsibilities
|
|
75
|
-
- Capabilities
|
|
76
|
-
- Interaction Protocol
|
|
77
|
-
- Mandatory Protocols
|
|
78
|
-
- Workflows
|
|
79
|
-
- Examples
|
|
80
|
-
|
|
81
|
-
**Recommended sections:**
|
|
82
|
-
- Integration Points
|
|
83
|
-
- Conflict Priorities
|
|
84
|
-
- Best Practices
|
|
85
|
-
- Advanced Features
|
|
86
|
-
- Error Handling
|
|
87
|
-
- Success Metrics
|
|
88
|
-
|
|
89
|
-
## Initial Agent
|
|
90
|
-
|
|
91
|
-
When auto-setup detects no agents exist, it creates:
|
|
92
|
-
|
|
93
|
-
**kiro-master** - Interactive Kiro feature management with CRUD operations for MCP servers, hooks, agents, specs, powers, and steering documents. Includes .kiro/ directory maintenance, steering optimization, refactoring, and comprehensive analysis capabilities
|
|
94
|
-
|
|
95
|
-
## Integration Notes
|
|
96
|
-
|
|
97
|
-
- Works alongside steering documents
|
|
98
|
-
- Compatible with chit-chat mode
|
|
99
|
-
- Preserves ADHD-C optimizations
|
|
100
|
-
- **Mode system** - Works seamlessly with modes (see `modes-system.md`)
|
|
101
|
-
- **Context preservation** - File changes and conversation history persist across agent switches
|
|
102
|
-
|
|
103
|
-
## System Notes
|
|
104
|
-
|
|
105
|
-
- Prototype system using Instruction Alias pattern
|
|
106
|
-
- Relies on AI understanding and following instructions
|
|
107
|
-
- May need iteration based on actual usage
|
|
108
|
-
|
|
109
|
-
---
|
|
110
|
-
|
|
111
|
-
**Agent system ready. Use `/agents` to begin or `/agents {name}` to activate specific agent.**
|
|
@@ -1,98 +0,0 @@
|
|
|
1
|
-
---
|
|
2
|
-
inclusion: always
|
|
3
|
-
description: Kiro mode switching system for flexible interaction styles (vibe/spec modes)
|
|
4
|
-
---
|
|
5
|
-
|
|
6
|
-
# Modes System
|
|
7
|
-
|
|
8
|
-
Mode switching system for Kiro that enables different interaction styles and workflows.
|
|
9
|
-
|
|
10
|
-
**For complete documentation, examples, and troubleshooting, see `docs/modes-system-guide.md`**
|
|
11
|
-
|
|
12
|
-
## Available Modes
|
|
13
|
-
|
|
14
|
-
- **vibe** - Flexible, conversational development assistance
|
|
15
|
-
- **spec** - Structured feature development with requirements, design, and tasks
|
|
16
|
-
|
|
17
|
-
## Mode Commands
|
|
18
|
-
|
|
19
|
-
### Command 1: Switch Kiro Mode
|
|
20
|
-
|
|
21
|
-
<alias>
|
|
22
|
-
<trigger>/modes {mode_name}</trigger>
|
|
23
|
-
<definition>
|
|
24
|
-
## Mode Switch: {mode_name}
|
|
25
|
-
|
|
26
|
-
You are now switching to **{mode_name} mode**.
|
|
27
|
-
|
|
28
|
-
**Load and execute mode switching protocol:**
|
|
29
|
-
1. Read `kiro-{mode_name}-mode.md` from agent-system directory into context
|
|
30
|
-
2. Read `~/.kiro/steering/kiro-agents/protocols/mode-switching.mdx` into context
|
|
31
|
-
3. Follow all steps from the "Mode Switch Steps" section in mode-switching.mdx
|
|
32
|
-
4. Use `{mode_name}` as the mode identifier throughout the protocol
|
|
33
|
-
</definition>
|
|
34
|
-
</alias>
|
|
35
|
-
|
|
36
|
-
### Command 2: Interactive Mode Management
|
|
37
|
-
|
|
38
|
-
For interactive mode management, use the `/modes` slash command (without parameters).
|
|
39
|
-
This provides a visual interface for:
|
|
40
|
-
- Viewing available modes
|
|
41
|
-
- Comparing modes
|
|
42
|
-
- Switching modes
|
|
43
|
-
- Getting help
|
|
44
|
-
|
|
45
|
-
**Usage:**
|
|
46
|
-
- Interactive: `/modes` (no parameters, loads `modes.md` steering file)
|
|
47
|
-
- Direct: `/modes {name}` (with mode name, uses alias above)
|
|
48
|
-
|
|
49
|
-
## Quick Reference
|
|
50
|
-
|
|
51
|
-
### Activation Protocol Summary
|
|
52
|
-
|
|
53
|
-
When `/modes {name}` is executed:
|
|
54
|
-
|
|
55
|
-
1. **Load** - Read mode definition and parse configuration
|
|
56
|
-
2. **Check** - If from spec mode, show warning and confirm
|
|
57
|
-
3. **Apply** - Follow mode protocols, set up context
|
|
58
|
-
4. **Begin** - Start interaction per mode's protocol
|
|
59
|
-
|
|
60
|
-
### Management Protocol Summary
|
|
61
|
-
|
|
62
|
-
When `/modes` is executed:
|
|
63
|
-
|
|
64
|
-
1. **Activate chit-chat** - Load chit-chat.md, use diff blocks
|
|
65
|
-
2. **Detect modes** - Scan for `kiro-*-mode.md` files
|
|
66
|
-
3. **Present choices** - Show modes, offer operations
|
|
67
|
-
4. **Handle operations** - Switching, viewing, comparison, help
|
|
68
|
-
|
|
69
|
-
## Mode and Agent Coordination
|
|
70
|
-
|
|
71
|
-
**Modes provide the framework:**
|
|
72
|
-
- Vibe mode: Flexible, conversational interaction
|
|
73
|
-
- Spec mode: Structured workflow with approval gates
|
|
74
|
-
|
|
75
|
-
**Agents provide specialized capabilities:**
|
|
76
|
-
- kiro-master: Feature management and .kiro/ operations
|
|
77
|
-
- Custom agents: Domain-specific expertise
|
|
78
|
-
|
|
79
|
-
**They work together:**
|
|
80
|
-
- Modes can activate agents: `/modes spec` then `/agents kiro-master`
|
|
81
|
-
- Agents inherit mode protocols: Agent in spec mode follows spec workflow
|
|
82
|
-
- Context preserved: Switch modes/agents without losing work
|
|
83
|
-
- Layered capabilities: Mode + Agent + Strict mode all active simultaneously
|
|
84
|
-
|
|
85
|
-
## System Notes
|
|
86
|
-
|
|
87
|
-
- Prototype system using Instruction Alias pattern
|
|
88
|
-
- Relies on AI understanding and following instructions
|
|
89
|
-
- Works seamlessly with agent system
|
|
90
|
-
|
|
91
|
-
---
|
|
92
|
-
|
|
93
|
-
**Modes system ready.**
|
|
94
|
-
|
|
95
|
-
**Quick start:**
|
|
96
|
-
- `/modes vibe` - Flexible development
|
|
97
|
-
- `/modes spec` - Structured planning
|
|
98
|
-
- `/modes` - Interactive mode management
|
|
File without changes
|
|
File without changes
|
|
File without changes
|