claude-flow-novice 2.14.14 β 2.14.16
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/.claude/commands/cfn-loop-cli.md +21 -27
- package/claude-assets/commands/cfn-loop-cli.md +21 -27
- package/claude-assets/skills/cfn-multi-coordinator-planning/map-dependencies-conflicts.sh +375 -375
- package/claude-assets/skills/cfn-multi-coordinator-planning/plan-coordinator-resources.sh +257 -257
- package/claude-assets/skills/cfn-multi-coordinator-planning/plan-multi-coordinator-work.sh +266 -266
- package/claude-assets/skills/cfn-multi-coordinator-planning/plan-risk-rollout.sh +349 -349
- package/claude-assets/skills/cfn-multi-coordinator-planning/test-multi-coordinator-planning.sh +337 -337
- package/claude-assets/skills/cfn-multi-coordinator-planning/validate-task-planning.sh +188 -188
- package/dist/agent/skill-mcp-selector.js +459 -0
- package/dist/agent/skill-mcp-selector.js.map +1 -0
- package/dist/cli/agent-token-manager.js +382 -0
- package/dist/cli/agent-token-manager.js.map +1 -0
- package/dist/cli/config-manager.js +91 -109
- package/dist/cli/config-manager.js.map +1 -1
- package/dist/mcp/auth-middleware.js +367 -0
- package/dist/mcp/auth-middleware.js.map +1 -0
- package/dist/mcp/playwright-mcp-server-auth.js +515 -0
- package/dist/mcp/playwright-mcp-server-auth.js.map +1 -0
- package/package.json +1 -1
|
@@ -79,14 +79,13 @@ Savings: 64% with custom routing, 95-98% vs all-Task
|
|
|
79
79
|
|
|
80
80
|
## π¨ CRITICAL EXECUTION INSTRUCTIONS
|
|
81
81
|
|
|
82
|
-
**YOU ARE MAIN CHAT. YOU SPAWN ONLY
|
|
82
|
+
**YOU ARE MAIN CHAT. YOU SPAWN ONLY THE cfn-v3-coordinator AGENT.**
|
|
83
83
|
|
|
84
84
|
**DO NOT spawn Loop 3 agents (backend-dev, researcher, etc.)**
|
|
85
85
|
**DO NOT spawn Loop 2 agents (reviewer, tester, etc.)**
|
|
86
86
|
**DO NOT spawn product-owner**
|
|
87
|
-
**DO NOT ask user which agents to use**
|
|
88
87
|
|
|
89
|
-
The coordinator handles ALL agent spawning internally via CLI.
|
|
88
|
+
The coordinator handles ALL agent spawning internally via enhanced CLI mode v3.0.
|
|
90
89
|
|
|
91
90
|
---
|
|
92
91
|
|
|
@@ -94,14 +93,14 @@ The coordinator handles ALL agent spawning internally via CLI.
|
|
|
94
93
|
|
|
95
94
|
```javascript
|
|
96
95
|
Task("cfn-v3-coordinator", `
|
|
97
|
-
CFN LOOP CLI MODE - PRODUCTION EXECUTION
|
|
96
|
+
CFN LOOP CLI MODE - ENHANCED PRODUCTION EXECUTION v3.0
|
|
98
97
|
|
|
99
98
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
100
99
|
TASK SPECIFICATION
|
|
101
100
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
102
101
|
|
|
103
102
|
Task Description: $ARGUMENTS
|
|
104
|
-
Task ID: cfn-cli-$(date +%s)
|
|
103
|
+
Task ID: cfn-cli-$(date +%s%N | tail -c 7)-${RANDOM}
|
|
105
104
|
Mode: ${mode.toUpperCase()}
|
|
106
105
|
|
|
107
106
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
@@ -126,7 +125,7 @@ Task("cfn-v3-coordinator", `
|
|
|
126
125
|
- Product Owner approval received
|
|
127
126
|
|
|
128
127
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
129
|
-
ORCHESTRATION CONFIGURATION
|
|
128
|
+
ENHANCED ORCHESTRATION CONFIGURATION v3.0
|
|
130
129
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
131
130
|
|
|
132
131
|
Mode: ${mode.toUpperCase()}
|
|
@@ -146,12 +145,12 @@ Task("cfn-v3-coordinator", `
|
|
|
146
145
|
Product Owner: product-owner
|
|
147
146
|
|
|
148
147
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
149
|
-
EXECUTION INSTRUCTIONS
|
|
148
|
+
ENHANCED EXECUTION INSTRUCTIONS v3.0
|
|
150
149
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
151
150
|
|
|
152
|
-
1. INVOKE ORCHESTRATOR (CLI spawning):
|
|
151
|
+
1. INVOKE ENHANCED ORCHESTRATOR (CLI spawning v3.0):
|
|
153
152
|
|
|
154
|
-
TASK_ID="cfn-cli-$(date +%s)"
|
|
153
|
+
TASK_ID="cfn-cli-$(date +%s%N | tail -c 7)-${RANDOM}"
|
|
155
154
|
MODE="${mode}"
|
|
156
155
|
LOOP3_AGENTS="backend-dev,researcher,devops" # Customize for task
|
|
157
156
|
LOOP2_AGENTS="reviewer,tester,architect,security-specialist" # Scale by complexity
|
|
@@ -164,8 +163,12 @@ Task("cfn-v3-coordinator", `
|
|
|
164
163
|
--product-owner "product-owner" \\
|
|
165
164
|
--max-iterations ${maxIterations}
|
|
166
165
|
|
|
167
|
-
2. ORCHESTRATOR HANDLES:
|
|
168
|
-
-
|
|
166
|
+
2. ENHANCED ORCHESTRATOR HANDLES v3.0:
|
|
167
|
+
- **Real-time monitoring** with automatic stuck agent recovery
|
|
168
|
+
- **Process health checking** and dead process cleanup
|
|
169
|
+
- **Protocol compliance** preventing "consensus on vapor" anti-patterns
|
|
170
|
+
- **Progress visibility** with detailed timestamped reports
|
|
171
|
+
- Spawns all agents via enhanced CLI (background)
|
|
169
172
|
- Loop 3: Gate check (β₯threshold) β PASS/ITERATE
|
|
170
173
|
- Loop 2: Consensus check (β₯threshold) β COMPLETE/ITERATE
|
|
171
174
|
- Product Owner: PROCEED/ITERATE/ABORT decision
|
|
@@ -190,28 +193,15 @@ Task("cfn-v3-coordinator", `
|
|
|
190
193
|
CRITICAL RULES
|
|
191
194
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
192
195
|
|
|
193
|
-
- DO NOT spawn agents with Task() - orchestrator uses CLI
|
|
196
|
+
- DO NOT spawn agents with Task() - orchestrator uses enhanced CLI
|
|
194
197
|
- ALL agents run in background via npx claude-flow-novice
|
|
195
198
|
- USE Redis BLPOP for loop dependencies
|
|
196
199
|
- AGENTS use Z.ai routing automatically (when enabled)
|
|
200
|
+
- ORCHESTRATOR provides real-time monitoring and recovery
|
|
197
201
|
- RETURN structured result when complete
|
|
198
202
|
`, "cfn-v3-coordinator")
|
|
199
203
|
```
|
|
200
204
|
|
|
201
|
-
## Autonomous Execution Rules
|
|
202
|
-
|
|
203
|
-
**YOU ARE FORBIDDEN FROM:**
|
|
204
|
-
- β Asking "Should I retry?" (coordinator handles automatically)
|
|
205
|
-
- β Asking "Proceed to consensus?" (orchestrator decides)
|
|
206
|
-
- β Waiting for approval during CFN Loop cycles
|
|
207
|
-
- β Spawning workers with Task() (use coordinator only)
|
|
208
|
-
|
|
209
|
-
**YOU MUST:**
|
|
210
|
-
- β
ALWAYS spawn single coordinator agent
|
|
211
|
-
- β
LET coordinator invoke orchestrator internally
|
|
212
|
-
- β
COORDINATOR handles all loop execution autonomously
|
|
213
|
-
- β
ONLY return to chat when complete or blocked
|
|
214
|
-
|
|
215
205
|
## CLI Mode Benefits
|
|
216
206
|
|
|
217
207
|
**Cost Savings:**
|
|
@@ -219,11 +209,15 @@ Task("cfn-v3-coordinator", `
|
|
|
219
209
|
- 95-98% savings vs Task tool spawning
|
|
220
210
|
- Scales linearly with iterations (Task mode scales exponentially)
|
|
221
211
|
|
|
222
|
-
**Production Features:**
|
|
212
|
+
**Production Features v3.0:**
|
|
223
213
|
- Background execution (no timeout issues)
|
|
224
214
|
- Redis state persistence (crash recovery)
|
|
225
215
|
- Zero-token waiting (BLPOP blocks without API calls)
|
|
226
216
|
- Web portal visibility (http://localhost:3000)
|
|
217
|
+
- **Enhanced monitoring**: Real-time agent progress tracking
|
|
218
|
+
- **Automatic recovery**: Dead process cleanup and agent restart
|
|
219
|
+
- **Protocol compliance**: Prevents "consensus on vapor" anti-patterns
|
|
220
|
+
- **Progress visibility**: Detailed reports with timestamps and health status
|
|
227
221
|
|
|
228
222
|
**Performance:**
|
|
229
223
|
- Parallel agent spawning (no sequential bottleneck)
|
|
@@ -79,14 +79,13 @@ Savings: 64% with custom routing, 95-98% vs all-Task
|
|
|
79
79
|
|
|
80
80
|
## π¨ CRITICAL EXECUTION INSTRUCTIONS
|
|
81
81
|
|
|
82
|
-
**YOU ARE MAIN CHAT. YOU SPAWN ONLY
|
|
82
|
+
**YOU ARE MAIN CHAT. YOU SPAWN ONLY THE cfn-v3-coordinator AGENT.**
|
|
83
83
|
|
|
84
84
|
**DO NOT spawn Loop 3 agents (backend-dev, researcher, etc.)**
|
|
85
85
|
**DO NOT spawn Loop 2 agents (reviewer, tester, etc.)**
|
|
86
86
|
**DO NOT spawn product-owner**
|
|
87
|
-
**DO NOT ask user which agents to use**
|
|
88
87
|
|
|
89
|
-
The coordinator handles ALL agent spawning internally via CLI.
|
|
88
|
+
The coordinator handles ALL agent spawning internally via enhanced CLI mode v3.0.
|
|
90
89
|
|
|
91
90
|
---
|
|
92
91
|
|
|
@@ -94,14 +93,14 @@ The coordinator handles ALL agent spawning internally via CLI.
|
|
|
94
93
|
|
|
95
94
|
```javascript
|
|
96
95
|
Task("cfn-v3-coordinator", `
|
|
97
|
-
CFN LOOP CLI MODE - PRODUCTION EXECUTION
|
|
96
|
+
CFN LOOP CLI MODE - ENHANCED PRODUCTION EXECUTION v3.0
|
|
98
97
|
|
|
99
98
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
100
99
|
TASK SPECIFICATION
|
|
101
100
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
102
101
|
|
|
103
102
|
Task Description: $ARGUMENTS
|
|
104
|
-
Task ID: cfn-cli-$(date +%s)
|
|
103
|
+
Task ID: cfn-cli-$(date +%s%N | tail -c 7)-${RANDOM}
|
|
105
104
|
Mode: ${mode.toUpperCase()}
|
|
106
105
|
|
|
107
106
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
@@ -126,7 +125,7 @@ Task("cfn-v3-coordinator", `
|
|
|
126
125
|
- Product Owner approval received
|
|
127
126
|
|
|
128
127
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
129
|
-
ORCHESTRATION CONFIGURATION
|
|
128
|
+
ENHANCED ORCHESTRATION CONFIGURATION v3.0
|
|
130
129
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
131
130
|
|
|
132
131
|
Mode: ${mode.toUpperCase()}
|
|
@@ -146,12 +145,12 @@ Task("cfn-v3-coordinator", `
|
|
|
146
145
|
Product Owner: product-owner
|
|
147
146
|
|
|
148
147
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
149
|
-
EXECUTION INSTRUCTIONS
|
|
148
|
+
ENHANCED EXECUTION INSTRUCTIONS v3.0
|
|
150
149
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
151
150
|
|
|
152
|
-
1. INVOKE ORCHESTRATOR (CLI spawning):
|
|
151
|
+
1. INVOKE ENHANCED ORCHESTRATOR (CLI spawning v3.0):
|
|
153
152
|
|
|
154
|
-
TASK_ID="cfn-cli-$(date +%s)"
|
|
153
|
+
TASK_ID="cfn-cli-$(date +%s%N | tail -c 7)-${RANDOM}"
|
|
155
154
|
MODE="${mode}"
|
|
156
155
|
LOOP3_AGENTS="backend-dev,researcher,devops" # Customize for task
|
|
157
156
|
LOOP2_AGENTS="reviewer,tester,architect,security-specialist" # Scale by complexity
|
|
@@ -164,8 +163,12 @@ Task("cfn-v3-coordinator", `
|
|
|
164
163
|
--product-owner "product-owner" \\
|
|
165
164
|
--max-iterations ${maxIterations}
|
|
166
165
|
|
|
167
|
-
2. ORCHESTRATOR HANDLES:
|
|
168
|
-
-
|
|
166
|
+
2. ENHANCED ORCHESTRATOR HANDLES v3.0:
|
|
167
|
+
- **Real-time monitoring** with automatic stuck agent recovery
|
|
168
|
+
- **Process health checking** and dead process cleanup
|
|
169
|
+
- **Protocol compliance** preventing "consensus on vapor" anti-patterns
|
|
170
|
+
- **Progress visibility** with detailed timestamped reports
|
|
171
|
+
- Spawns all agents via enhanced CLI (background)
|
|
169
172
|
- Loop 3: Gate check (β₯threshold) β PASS/ITERATE
|
|
170
173
|
- Loop 2: Consensus check (β₯threshold) β COMPLETE/ITERATE
|
|
171
174
|
- Product Owner: PROCEED/ITERATE/ABORT decision
|
|
@@ -190,28 +193,15 @@ Task("cfn-v3-coordinator", `
|
|
|
190
193
|
CRITICAL RULES
|
|
191
194
|
ββββββββββββββββββββββββββββββββββββββββββββββ
|
|
192
195
|
|
|
193
|
-
- DO NOT spawn agents with Task() - orchestrator uses CLI
|
|
196
|
+
- DO NOT spawn agents with Task() - orchestrator uses enhanced CLI
|
|
194
197
|
- ALL agents run in background via npx claude-flow-novice
|
|
195
198
|
- USE Redis BLPOP for loop dependencies
|
|
196
199
|
- AGENTS use Z.ai routing automatically (when enabled)
|
|
200
|
+
- ORCHESTRATOR provides real-time monitoring and recovery
|
|
197
201
|
- RETURN structured result when complete
|
|
198
202
|
`, "cfn-v3-coordinator")
|
|
199
203
|
```
|
|
200
204
|
|
|
201
|
-
## Autonomous Execution Rules
|
|
202
|
-
|
|
203
|
-
**YOU ARE FORBIDDEN FROM:**
|
|
204
|
-
- β Asking "Should I retry?" (coordinator handles automatically)
|
|
205
|
-
- β Asking "Proceed to consensus?" (orchestrator decides)
|
|
206
|
-
- β Waiting for approval during CFN Loop cycles
|
|
207
|
-
- β Spawning workers with Task() (use coordinator only)
|
|
208
|
-
|
|
209
|
-
**YOU MUST:**
|
|
210
|
-
- β
ALWAYS spawn single coordinator agent
|
|
211
|
-
- β
LET coordinator invoke orchestrator internally
|
|
212
|
-
- β
COORDINATOR handles all loop execution autonomously
|
|
213
|
-
- β
ONLY return to chat when complete or blocked
|
|
214
|
-
|
|
215
205
|
## CLI Mode Benefits
|
|
216
206
|
|
|
217
207
|
**Cost Savings:**
|
|
@@ -219,11 +209,15 @@ Task("cfn-v3-coordinator", `
|
|
|
219
209
|
- 95-98% savings vs Task tool spawning
|
|
220
210
|
- Scales linearly with iterations (Task mode scales exponentially)
|
|
221
211
|
|
|
222
|
-
**Production Features:**
|
|
212
|
+
**Production Features v3.0:**
|
|
223
213
|
- Background execution (no timeout issues)
|
|
224
214
|
- Redis state persistence (crash recovery)
|
|
225
215
|
- Zero-token waiting (BLPOP blocks without API calls)
|
|
226
216
|
- Web portal visibility (http://localhost:3000)
|
|
217
|
+
- **Enhanced monitoring**: Real-time agent progress tracking
|
|
218
|
+
- **Automatic recovery**: Dead process cleanup and agent restart
|
|
219
|
+
- **Protocol compliance**: Prevents "consensus on vapor" anti-patterns
|
|
220
|
+
- **Progress visibility**: Detailed reports with timestamps and health status
|
|
227
221
|
|
|
228
222
|
**Performance:**
|
|
229
223
|
- Parallel agent spawning (no sequential bottleneck)
|