groove-dev 0.26.19 → 0.26.20

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 (3) hide show
  1. package/node_modules/@groove-dev/daemon/src/process.js +7 -3
  2. package/package.json +1 -1
  3. package/packages/daemon/src/process.js +7 -3
package/node_modules/@groove-dev/daemon/src/process.js CHANGED
@@ -86,11 +86,13 @@ Focus on:
86
86
  - Identifying approaches and trade-offs
87
87
  - Writing structured plans with agent assignments
88
88
 
89
- After completing your plan, you MUST do two things:
89
+ After completing your plan, you MUST do two things — EVERY TIME, no exceptions:
90
90
 
91
91
  1. Write your team recommendation as a clear summary in your output so the user can review it.
92
92
 
93
- 2. Save a machine-readable team config to .groove/recommended-team.json using this EXACT format:
93
+ 2. Save a machine-readable team config to .groove/recommended-team.json using this EXACT format.
94
+ ALWAYS write this file, even if the user says "no task yet" or "just building the team."
95
+ For team-building without a specific task, use empty prompts ("prompt": "") — the agents will wait for instructions.
94
96
 
95
97
  For NEW projects (building something from scratch):
96
98
  {
@@ -126,10 +128,12 @@ MANDATORY RULES — NEVER SKIP THESE:
126
128
 
127
129
  3. Do NOT tell any agent to "wait for" another agent. Phase 2 handles sequencing automatically.
128
130
 
129
- 4. Set appropriate scopes. Write detailed prompts so each agent knows exactly what to build.
131
+ 4. Set appropriate scopes. If the user gave a specific task, write detailed prompts. If no task was given, use empty prompts ("prompt": "") — the agents will await instructions.
130
132
 
131
133
  5. NEVER instruct any agent to delete files from other projects or clean up unrelated code. Each agent must ONLY create and modify files relevant to its assigned tasks.
132
134
 
135
+ 6. You MUST always write .groove/recommended-team.json. NEVER skip it. The GUI depends on this file to show the Launch Team modal.
136
+
133
137
  IMPORTANT: Do not use markdown formatting like ** or ### in your output. Write in plain text with clean formatting. Use line breaks, dashes, and indentation for structure.
134
138
 
135
139
  `,
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "groove-dev",
3
- "version": "0.26.19",
3
+ "version": "0.26.20",
4
4
  "description": "Open-source agent orchestration layer — the AI company OS. Local model agent engine (GGUF/Ollama/llama-server), HuggingFace model browser, MCP integrations (Slack, Gmail, Stripe, 15+), agent scheduling (cron), business roles (CMO, CFO, EA). GUI dashboard, multi-agent coordination, zero cold-start, infinite sessions. Works with Claude Code, Codex, Gemini CLI, Ollama, any local model.",
5
5
  "license": "FSL-1.1-Apache-2.0",
6
6
  "author": "Groove Dev <hello@groovedev.ai> (https://groovedev.ai)",
package/packages/daemon/src/process.js CHANGED
@@ -86,11 +86,13 @@ Focus on:
86
86
  - Identifying approaches and trade-offs
87
87
  - Writing structured plans with agent assignments
88
88
 
89
- After completing your plan, you MUST do two things:
89
+ After completing your plan, you MUST do two things — EVERY TIME, no exceptions:
90
90
 
91
91
  1. Write your team recommendation as a clear summary in your output so the user can review it.
92
92
 
93
- 2. Save a machine-readable team config to .groove/recommended-team.json using this EXACT format:
93
+ 2. Save a machine-readable team config to .groove/recommended-team.json using this EXACT format.
94
+ ALWAYS write this file, even if the user says "no task yet" or "just building the team."
95
+ For team-building without a specific task, use empty prompts ("prompt": "") — the agents will wait for instructions.
94
96
 
95
97
  For NEW projects (building something from scratch):
96
98
  {
@@ -126,10 +128,12 @@ MANDATORY RULES — NEVER SKIP THESE:
126
128
 
127
129
  3. Do NOT tell any agent to "wait for" another agent. Phase 2 handles sequencing automatically.
128
130
 
129
- 4. Set appropriate scopes. Write detailed prompts so each agent knows exactly what to build.
131
+ 4. Set appropriate scopes. If the user gave a specific task, write detailed prompts. If no task was given, use empty prompts ("prompt": "") — the agents will await instructions.
130
132
 
131
133
  5. NEVER instruct any agent to delete files from other projects or clean up unrelated code. Each agent must ONLY create and modify files relevant to its assigned tasks.
132
134
 
135
+ 6. You MUST always write .groove/recommended-team.json. NEVER skip it. The GUI depends on this file to show the Launch Team modal.
136
+
133
137
  IMPORTANT: Do not use markdown formatting like ** or ### in your output. Write in plain text with clean formatting. Use line breaks, dashes, and indentation for structure.
134
138
 
135
139
  `,