bmm-opencode 1.4.0 โ 1.4.2
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.
|
@@ -286,19 +286,27 @@ Provide your expert perspective on this topic. Think deeply and thoroughly.
|
|
|
286
286
|
```
|
|
287
287
|
๐ **Meeting in Progress - Agents Thinking...**
|
|
288
288
|
|
|
289
|
-
| Agent | Model | Status | Task ID |
|
|
290
|
-
|
|
291
|
-
| ๐๏ธ Winston | o3 | ๐ analyzing... | bg_xxx |
|
|
292
|
-
| ๐ป Amelia | sonnet-4 | ๐ analyzing... | bg_yyy |
|
|
293
|
-
| ๐งช Murat | gemini-2.5-pro | ๐ analyzing... | bg_zzz |
|
|
289
|
+
| Agent | Model | Status | Task ID | Session ID |
|
|
290
|
+
|-------|-------|--------|---------|------------|
|
|
291
|
+
| ๐๏ธ Winston | o3 | ๐ analyzing... | `bg_xxx` | `ses_xxx` |
|
|
292
|
+
| ๐ป Amelia | sonnet-4 | ๐ analyzing... | `bg_yyy` | `ses_yyy` |
|
|
293
|
+
| ๐งช Murat | gemini-2.5-pro | ๐ analyzing... | `bg_zzz` | `ses_zzz` |
|
|
294
294
|
|
|
295
295
|
โณ Collecting insights from all participants...
|
|
296
296
|
|
|
297
|
+
### ๐ How to Check Progress
|
|
298
|
+
|
|
299
|
+
View any agent's real-time thinking:
|
|
300
|
+
```
|
|
301
|
+
background_output(task_id="<task_id>")
|
|
302
|
+
```
|
|
303
|
+
|
|
297
304
|
---
|
|
298
305
|
|
|
299
306
|
๐ก **While waiting, you can:**
|
|
300
307
|
- `+agent` - Add another agent to this meeting
|
|
301
308
|
- `status` - Check progress
|
|
309
|
+
- `background_output(task_id="bg_xxx")` - View specific agent's progress
|
|
302
310
|
- `cancel` - Cancel and restart
|
|
303
311
|
```
|
|
304
312
|
|
|
@@ -351,8 +359,16 @@ Provide your expert perspective on this topic. Think deeply and thoroughly.
|
|
|
351
359
|
| ๐ด High | [action] | architect, dev |
|
|
352
360
|
| ๐ก Medium | [action] | tea |
|
|
353
361
|
|
|
354
|
-
**โ
|
|
355
|
-
|
|
362
|
+
**โ Decisions Needed:**
|
|
363
|
+
|
|
364
|
+
When the synthesis reveals open questions or divergent views requiring user decision, present numbered options:
|
|
365
|
+
|
|
366
|
+
| # | Decision | Options |
|
|
367
|
+
|---|----------|---------|
|
|
368
|
+
| 1 | [decision topic] | [A] option-a / [B] option-b / [C] option-c |
|
|
369
|
+
| 2 | [decision topic] | [A] option-a / [B] option-b |
|
|
370
|
+
|
|
371
|
+
> Reply with the number + letter (e.g., `1A`, `2B`) to decide, or ask agents to elaborate.
|
|
356
372
|
|
|
357
373
|
**๐ Action Items:**
|
|
358
374
|
- [ ] [item 1]
|
|
@@ -365,6 +381,7 @@ Provide your expert perspective on this topic. Think deeply and thoroughly.
|
|
|
365
381
|
- `@agent question` - Ask specific agent a follow-up
|
|
366
382
|
- `deep [topic]` - Deep dive on specific topic
|
|
367
383
|
- `vote [options]` - Get agent votes on decision
|
|
384
|
+
- `decide [#]` - Revisit a specific decision point
|
|
368
385
|
- `summarize` - Get executive summary
|
|
369
386
|
- `report` - End meeting and generate summary report file (choose summary or detailed)
|
|
370
387
|
- `exit` - End meeting
|
|
@@ -476,6 +493,7 @@ meeting_state:
|
|
|
476
493
|
| `@all [question]` | Ask all agents | `@all thoughts on timeline?` |
|
|
477
494
|
| `deep [topic]` | Deep dive | `deep security concerns` |
|
|
478
495
|
| `vote [question]` | Get agent votes | `vote monolith vs microservice` |
|
|
496
|
+
| `decide [#]` | Revisit decision point | `decide 1` |
|
|
479
497
|
| `status` | Show meeting status | |
|
|
480
498
|
| `list` | Show all available agents | |
|
|
481
499
|
| `attendees` | Show current participants | |
|
|
@@ -785,6 +803,7 @@ Thank you for using Party Mode! ๐
|
|
|
785
803
|
11. **GRACEFUL EXIT** - Summarize meeting when ending
|
|
786
804
|
12. **HANDLE ERRORS** - Provide helpful error messages
|
|
787
805
|
13. **ALWAYS GENERATE TITLES** - Every assigned agent MUST have a Title = `AgentName [model]: [first impression]`. The model is the AI model powering that agent (determined by the `delegate_task` category). The first impression is a short, punchy one-liner (โค15 words) reflecting the agent's instinctive gut reaction to the meeting topic from their area of expertise. Generate titles in Phase 2 (recommendation) and carry them through Phase 3 (execution) and Phase 4 (results). Late-arriving agents (`+agent`) also get a title upon joining.
|
|
806
|
+
14. **PRESENT OPTIONS FOR DECISIONS** - When synthesis reveals divergent views, open questions, or any point requiring user decision, ALWAYS present numbered options with clear choices (e.g., `1A`, `1B`). Never leave decisions as open-ended text โ structure them as actionable choices the user can select from.
|
|
788
807
|
|
|
789
808
|
---
|
|
790
809
|
|
|
@@ -229,6 +229,18 @@ delegate_task({
|
|
|
229
229
|
|
|
230
230
|
---
|
|
231
231
|
|
|
232
|
+
## MONITORING BACKGROUND AGENTS
|
|
233
|
+
|
|
234
|
+
When agents are running in background, the workflow displays a progress table with **Task ID** and **Session ID** for each agent. Users can check any agent's real-time progress:
|
|
235
|
+
|
|
236
|
+
```
|
|
237
|
+
background_output(task_id="<task_id>")
|
|
238
|
+
```
|
|
239
|
+
|
|
240
|
+
This returns the agent's current status, latest tool calls, and output so far โ allowing you to monitor what each dev-story or code-review agent is doing without waiting for completion.
|
|
241
|
+
|
|
242
|
+
---
|
|
243
|
+
|
|
232
244
|
## STATE SCHEMA
|
|
233
245
|
|
|
234
246
|
```yaml
|
|
@@ -185,6 +185,23 @@ When an agent asks the user a specific question:
|
|
|
185
185
|
- Clearly highlight the questioning agent and their question
|
|
186
186
|
- Wait for user response before any agent continues
|
|
187
187
|
|
|
188
|
+
### Decision Points
|
|
189
|
+
|
|
190
|
+
When agents identify divergent views, trade-offs, or questions requiring user decision, ALWAYS present structured numbered options:
|
|
191
|
+
|
|
192
|
+
```
|
|
193
|
+
โ **Decisions Needed**
|
|
194
|
+
|
|
195
|
+
| # | Decision | Options |
|
|
196
|
+
|---|----------|---------|
|
|
197
|
+
| 1 | [topic] | [A] option-a / [B] option-b / [C] option-c |
|
|
198
|
+
| 2 | [topic] | [A] option-a / [B] option-b |
|
|
199
|
+
|
|
200
|
+
> Reply with number + letter (e.g., `1A`, `2B`) to decide, or ask agents to elaborate.
|
|
201
|
+
```
|
|
202
|
+
|
|
203
|
+
Never leave decisions as open-ended text โ structure them as actionable choices.
|
|
204
|
+
|
|
188
205
|
### Inter-Agent Questions
|
|
189
206
|
|
|
190
207
|
Agents can question each other and respond naturally within the same round for dynamic conversation.
|
package/README.md
CHANGED
|
@@ -163,6 +163,8 @@ curl -s https://raw.githubusercontent.com/Jack-R-Hong/BMM-opencode/refs/heads/ma
|
|
|
163
163
|
|
|
164
164
|
| Version | Agents | Skills | Changes |
|
|
165
165
|
|---------|--------|--------|---------|
|
|
166
|
+
| 1.4.2 | 19 | 62 | Party-mode: structured decision options + background agent progress monitoring |
|
|
167
|
+
| 1.4.1 | 19 | 62 | Dev-team-mode: show session_id and progress monitoring for background agents |
|
|
166
168
|
| 1.4.0 | 19 | 62 | Add per-agent LLM model configuration with multi-provider alternatives |
|
|
167
169
|
| 1.3.2 | 19 | 62 | Dev-team-mode: delegated mode for orchestrator support |
|
|
168
170
|
| 1.3.1 | 19 | 62 | Party-mode: agent titles with first impressions |
|
package/package.json
CHANGED