convoke-agents 2.3.0 → 2.4.0

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 (30) hide show
  1. package/CHANGELOG.md +36 -0
  2. package/INSTALLATION.md +1 -1
  3. package/README.md +90 -15
  4. package/_bmad/bme/_enhance/config.yaml +8 -0
  5. package/_bmad/bme/_enhance/extensions/bmm-pm.yaml +9 -0
  6. package/_bmad/bme/_enhance/guides/.gitkeep +0 -0
  7. package/_bmad/bme/_enhance/guides/ENHANCE-GUIDE.md +252 -0
  8. package/_bmad/bme/_enhance/workflows/initiatives-backlog/SKILL.md +6 -0
  9. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/.gitkeep +0 -0
  10. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-01-init.md +106 -0
  11. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-02-gather.md +136 -0
  12. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-03-score.md +146 -0
  13. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-04-prioritize.md +181 -0
  14. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-r/.gitkeep +0 -0
  15. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-r/step-r-01-load.md +120 -0
  16. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-r/step-r-02-rescore.md +141 -0
  17. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-r/step-r-03-update.md +154 -0
  18. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-t/.gitkeep +0 -0
  19. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-t/step-t-01-ingest.md +86 -0
  20. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-t/step-t-02-extract.md +169 -0
  21. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-t/step-t-03-score.md +147 -0
  22. package/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-t/step-t-04-update.md +155 -0
  23. package/_bmad/bme/_enhance/workflows/initiatives-backlog/templates/backlog-format-spec.md +219 -0
  24. package/_bmad/bme/_enhance/workflows/initiatives-backlog/templates/rice-scoring-guide.md +154 -0
  25. package/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md +88 -0
  26. package/_bmad/bme/_vortex/config.yaml +1 -1
  27. package/package.json +2 -1
  28. package/scripts/update/lib/refresh-installation.js +139 -0
  29. package/scripts/update/lib/validator.js +122 -1
  30. package/scripts/update/migrations/registry.js +58 -20
package/CHANGELOG.md CHANGED
@@ -7,6 +7,42 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
7
7
 
8
8
  ---
9
9
 
10
+ ## [2.4.0] - 2026-03-15
11
+
12
+ ### Added
13
+
14
+ - **Enhance module — Agent Skills architecture** — New module type alongside Teams. Skills add capabilities to existing agents via menu patching without modifying agent files.
15
+ - **Initiatives Backlog skill (PM agent)** — RICE-scored backlog management with 3 modes:
16
+ - **Triage** — Ingest review findings, extract actionable items, propose RICE scores with two-gate validation, append to existing backlog (4 step files)
17
+ - **Review** — Walk through existing items one at a time, rescore where priorities shifted, regenerate prioritized view (3 step files)
18
+ - **Create** — Gather initiatives interactively, batch-score with RICE, generate complete backlog from scratch (4 step files)
19
+ - **Enhance module installer integration** — File copy, PM agent menu patching (`[IB] Initiatives Backlog`), 6-point verification in `convoke-doctor`, skill wrapper + manifest registration
20
+ - **ENHANCE-GUIDE.md** — Pattern documentation for module authors: directory structure, step file architecture, menu patching, config registration, verification integration
21
+ - **Enhance module validation** — `validateEnhanceModule()` with 6-point check: directory exists, entry point resolves, menu patch present, config valid, filesystem consistency, skill wrapper exists
22
+
23
+ ### Changed
24
+
25
+ - **README** — Repositioned Convoke around two extensibility axes: Teams (new agents) and Skills (new capabilities). Full Enhance section with mode diagram, activation instructions, and link to pattern guide. Updated architecture diagram, install tree, and roadmap.
26
+
27
+ ---
28
+
29
+ ## [2.3.1] - 2026-03-15
30
+
31
+ ### Fixed
32
+
33
+ - **Migration chaining bug** — `getMigrationsFor()` now walks the full migration chain instead of only matching the first hop. Users jumping multiple versions (e.g., `1.0.x` → `2.x`) previously skipped all intermediate migrations silently. The `1.5.x-to-1.6.0` config delta (Wave 3 agents) was being skipped for anyone upgrading from `1.0.x–1.4.x`.
34
+ - **README version badge** — Updated from stale `2.0.1` to current version
35
+ - **INSTALLATION.md** — Added missing `-p convoke-agents` to `npx` command in troubleshooting section
36
+ - **BMAD-METHOD-COMPATIBILITY.md** — Fixed wrong package name (`convoke` → `convoke-agents`) and added missing `-p convoke-agents` to all `npx` commands
37
+ - **Archived outdated docs** — PUBLISHING-GUIDE.md and TEST-PLAN-REAL-INSTALL.md marked as historical (contained legacy commands)
38
+
39
+ ### Added
40
+
41
+ - **Chain traversal tests** — 10 new unit tests for migration chain resolution, parallel entry exclusion, and breaking change detection across chains
42
+ - **Multi-version integration test** — Verifies full `1.5.2 → current` update path with all 3 deltas executing and complete migration history
43
+
44
+ ---
45
+
10
46
  ## [2.2.0] - 2026-03-14
11
47
 
12
48
  ### Changed
package/INSTALLATION.md CHANGED
@@ -181,7 +181,7 @@ The installer preserves your custom settings and only adds missing entries. To f
181
181
 
182
182
  ```bash
183
183
  rm -rf _bmad/bme/_vortex/
184
- npx convoke-install-vortex
184
+ npx -p convoke-agents convoke-install-vortex
185
185
  ```
186
186
 
187
187
  ### Installation succeeds but agents don't activate
package/README.md CHANGED
@@ -10,12 +10,14 @@
10
10
  Agent teams for complex systems
11
11
  ```
12
12
 
13
- [![Version](https://img.shields.io/badge/version-2.0.1-blue)](https://github.com/amalik/convoke-agents)
13
+ [![Version](https://img.shields.io/badge/version-2.4.0-blue)](https://github.com/amalik/convoke-agents)
14
14
  [![License](https://img.shields.io/badge/license-MIT-blue)](LICENSE)
15
15
 
16
16
  </div>
17
17
 
18
- Convoke organizes AI agents into domain-specialized teams. Each team brings deep expertise to a specific discipline — product discovery, implementation, operations, or whatever your domain requires. **Vortex**, the product discovery team, is the first. You can install teams independently or combine them.
18
+ Convoke extends AI agents with two types of installable modules: **Teams** bring new agents for a domain, **Skills** add new capabilities to existing agents. Both are top-level modules install them independently or combine them.
19
+
20
+ **Vortex** is the first team (product discovery, 7 agents). **Enhance** is the first skill (RICE backlog management for the PM agent). More of each are coming.
19
21
 
20
22
  ---
21
23
 
@@ -200,14 +202,20 @@ Each workflow ends with a Compass routing suggestion. You don't need to follow a
200
202
 
201
203
  ```
202
204
  your-project/
203
- ├── _bmad/bme/_vortex/
204
- │ ├── agents/ # 7 agent definition files
205
- │ ├── workflows/ # 22 workflows
206
- │ ├── contracts/ # Artifact contract schemas
207
- │ ├── guides/ # User guides (all 7 agents)
208
- └── config.yaml # Configuration
205
+ ├── _bmad/bme/
206
+ │ ├── _vortex/ # Team: Product Discovery
207
+ ├── agents/ # 7 agent definition files
208
+ ├── workflows/ # 22 workflows
209
+ ├── contracts/ # Artifact contract schemas
210
+ │ ├── guides/ # User guides (all 7 agents)
211
+ │ │ └── config.yaml # Configuration
212
+ │ └── _enhance/ # Skill: Agent Capability Upgrades
213
+ │ ├── workflows/ # Skill workflows (initiatives-backlog)
214
+ │ ├── extensions/ # Agent menu patch descriptors
215
+ │ ├── guides/ # Module author guide
216
+ │ └── config.yaml # Configuration
209
217
  └── _bmad-output/
210
- └── vortex-artifacts/ # Generated artifacts
218
+ └── vortex-artifacts/ # Generated artifacts
211
219
  ```
212
220
 
213
221
  ---
@@ -256,21 +264,87 @@ For detailed workflow descriptions and usage examples, see the [Agent Guide](doc
256
264
 
257
265
  ---
258
266
 
267
+ ## Enhance — Agent Skills
268
+
269
+ **Add new capabilities to existing agents without modifying them**
270
+
271
+ [![Workflows](https://img.shields.io/badge/workflows-1-success)](_bmad/bme/_enhance/guides/ENHANCE-GUIDE.md)
272
+ [![Modes](https://img.shields.io/badge/modes-3-blue)](_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md)
273
+
274
+ Skills are the other half of Convoke's extensibility. While Teams bring new agents, Skills give existing agents new workflows — installed via menu patching, not agent modification.
275
+
276
+ ### Initiatives Backlog (PM Agent)
277
+
278
+ The first Enhance skill adds RICE-scored backlog management to the PM agent. Three modes cover the full lifecycle:
279
+
280
+ ```
281
+ Initiatives Backlog
282
+
283
+ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐
284
+ │ [T] Triage │ │ [R] Review │ │ [C] Create │
285
+ │ Ingest new │ │ Rescore │ │ Bootstrap │
286
+ │ findings │ │ existing │ │ from scratch│
287
+ └──────┬──────┘ └──────┬──────┘ └──────┬──────┘
288
+ │ │ │
289
+ ▼ ▼ ▼
290
+ ┌──────────────────────────────────────────────────┐
291
+ │ initiatives-backlog.md │
292
+ │ RICE-scored · Categorized · Change-tracked │
293
+ └──────────────────────────────────────────────────┘
294
+ ```
295
+
296
+ | Mode | What it does |
297
+ |------|-------------|
298
+ | **Triage** | Ingest review findings, extract actionable items, propose RICE scores with two-gate validation, append to existing backlog |
299
+ | **Review** | Walk through items one at a time, rescore where priorities have shifted, regenerate prioritized view |
300
+ | **Create** | Gather initiatives interactively, batch-score with RICE, generate a complete backlog from scratch |
301
+
302
+ #### Activate
303
+
304
+ The Enhance skill appears in the PM agent's menu after installation:
305
+
306
+ ```
307
+ [IB] Initiatives Backlog (Convoke Enhance)
308
+ ```
309
+
310
+ Select it from the PM agent menu, or activate directly:
311
+
312
+ **Claude Code (skills)**
313
+ ```
314
+ /bmad-enhance-initiatives-backlog
315
+ ```
316
+
317
+ **Claude Code (terminal) / Other AI assistants**
318
+ ```bash
319
+ cat _bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md
320
+ ```
321
+
322
+ ### Building Your Own Skills
323
+
324
+ The [Enhance Guide](_bmad/bme/_enhance/guides/ENHANCE-GUIDE.md) documents the complete pattern for creating new skills: directory structure, step file architecture, agent menu patching, config registration, and verification integration. It uses the initiatives-backlog skill as the canonical example throughout.
325
+
326
+ Max 2-3 skills per agent to prevent menu bloat and maintain agent focus.
327
+
328
+ ---
329
+
259
330
  ## How It Fits with BMAD Core
260
331
 
261
332
  Convoke handles **discovery and validation**. BMAD Core handles **implementation**.
262
333
 
263
334
  ```
264
- Convoke Teams BMAD Core
335
+ Convoke Modules BMAD Core
265
336
  ┌──────────────────────────────┐ ┌──────────────────────┐
266
- Vortex (Product Discovery) ──────> PM → Architect → Dev
267
- "Should we build this?" "Let's build it"
337
+ Teams
338
+ Vortex (Product Discovery) ──────> PM Architect → Dev
339
+ │ [Future teams] │ │ "Let's build it" │
268
340
  │ │ <────── │ │
269
- [Future teams] │ signals │ │
341
+ Skills │ signals │ │
342
+ │ Enhance (Agent Upgrades) │ ──────> │ │
343
+ │ [Future skills] │ │ │
270
344
  └──────────────────────────────┘ └──────────────────────┘
271
345
  ```
272
346
 
273
- Convoke works standalone or as an extension — no BMAD Method installation required.
347
+ Teams and Skills are peer module types — both installable, both independent. Convoke works standalone or as an extension — no BMAD Method installation required.
274
348
 
275
349
  ---
276
350
 
@@ -293,7 +367,8 @@ Convoke works standalone or as an extension — no BMAD Method installation requ
293
367
  - **v1.6.x** — Wave 3: Complete 7-stream Vortex (added Mila, Liam, Noah — 7 agents, 22 workflows, handoff contracts, Compass routing)
294
368
  - **v1.7.0** — Wave 4: Quality & onboarding (P0 test suite, docs audit tool, all 22 workflows production-ready, README overhaul, package size fix)
295
369
  - **v2.0.0** — Product renamed to Convoke. CLI commands renamed to `convoke-*`. Package: `npm install convoke-agents`
296
- - **Next** — Additional domain-specialized teams, multi-agent collaboration, cross-team workflows
370
+ - **v2.x** — Enhance module: Skills architecture, RICE initiatives-backlog skill (Triage/Review/Create modes), module author pattern guide
371
+ - **Next** — Additional teams, additional skills, multi-agent collaboration, cross-module workflows
297
372
 
298
373
  ---
299
374
 
@@ -0,0 +1,8 @@
1
+ name: enhance
2
+ version: 1.0.0
3
+ description: "Enhance module — capability upgrades for existing BMAD agents"
4
+ workflows:
5
+ - name: initiatives-backlog
6
+ entry: workflows/initiatives-backlog/workflow.md
7
+ target_agent: bmm/agents/pm.md
8
+ menu_patch_name: "initiatives-backlog"
@@ -0,0 +1,9 @@
1
+ # Extension descriptor for John PM agent
2
+ # v1: Documentation-only — describes the patch-based menu item added by the installer
3
+ # v3: Will be read dynamically by agents via <extensions> tag (future BMAD Method proposal)
4
+
5
+ target_agent: bmm/pm
6
+ menu_items:
7
+ - cmd: "IB or fuzzy match on initiatives-backlog"
8
+ exec: "{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md"
9
+ label: "[IB] 📦 Initiatives Backlog (Convoke Enhance)"
File without changes
@@ -0,0 +1,252 @@
1
+ # Enhance Module — Pattern Guide for Module Authors
2
+
3
+ Create new enhancement workflows for existing BMAD agents by following the patterns established in the initiatives-backlog workflow. This guide covers directory layout, step file architecture, agent menu patching, config registration, and verification.
4
+
5
+ ---
6
+
7
+ ## Directory Structure
8
+
9
+ Every enhancement workflow lives under `_bmad/bme/_enhance/workflows/[name]/`:
10
+
11
+ ```
12
+ _bmad/bme/_enhance/
13
+ ├── config.yaml # Module config — register all workflows here
14
+ ├── extensions/
15
+ │ └── bmm-[agent].yaml # Extension descriptor (v1: docs-only)
16
+ ├── workflows/
17
+ │ └── [workflow-name]/
18
+ │ ├── workflow.md # Entry point — mode menu + dispatch
19
+ │ ├── SKILL.md # Skill wrapper (auto-generated by installer)
20
+ │ ├── steps-[mode]/ # One folder per mode
21
+ │ │ ├── step-[mode]-01-[name].md
22
+ │ │ ├── step-[mode]-02-[name].md
23
+ │ │ └── ...
24
+ │ └── templates/ # Reference docs loaded by steps at runtime
25
+ │ ├── [template-1].md
26
+ │ └── [template-2].md
27
+ └── guides/
28
+ └── ENHANCE-GUIDE.md # This file
29
+ ```
30
+
31
+ **Canonical example:** `workflows/initiatives-backlog/` — tri-modal workflow (Triage/Review/Create) with 11 step files across 3 mode folders and 2 templates.
32
+
33
+ **Naming conventions:**
34
+ - Workflow folders: kebab-case (`initiatives-backlog`, `sprint-health`)
35
+ - Step folders: `steps-[mode-letter]/` (e.g., `steps-t/`, `steps-r/`, `steps-c/`)
36
+ - Step files: `step-[mode]-[NN]-[name].md` (e.g., `step-t-01-ingest.md`)
37
+ - Templates: descriptive kebab-case (e.g., `rice-scoring-guide.md`, `backlog-format-spec.md`)
38
+
39
+ ---
40
+
41
+ ## Workflow Entry Point
42
+
43
+ `workflow.md` is the entry point loaded when the user selects the enhancement from the agent menu. It should:
44
+
45
+ 1. **Load config** from `_bmad/bme/_enhance/config.yaml`
46
+ 2. **Present a mode menu** (if multi-modal) or proceed directly (if single-mode)
47
+ 3. **Dispatch to the first step file** based on user selection
48
+ 4. **Define return-to-menu convention** — final steps reload `workflow.md` to re-present the menu
49
+
50
+ See `workflows/initiatives-backlog/workflow.md` for the canonical entry point pattern.
51
+
52
+ ---
53
+
54
+ ## Step File Architecture
55
+
56
+ Every step file follows the BMAD standard structure:
57
+
58
+ ```markdown
59
+ ---
60
+ name: 'step-[mode]-[NN]-[name]'
61
+ description: '[one-line description]'
62
+ nextStepFile: '{project-root}/_bmad/bme/_enhance/workflows/[name]/steps-[mode]/step-[mode]-[NN+1]-[name].md'
63
+ outputFile: '{planning_artifacts}/[output-file].md'
64
+ templateFile: '{project-root}/_bmad/bme/_enhance/workflows/[name]/templates/[template].md'
65
+ ---
66
+
67
+ # Step N: [Title]
68
+
69
+ ## STEP GOAL:
70
+ ## MANDATORY EXECUTION RULES (READ FIRST):
71
+ ### Universal Rules:
72
+ ### Role Reinforcement:
73
+ ### Step-Specific Rules:
74
+ ## EXECUTION PROTOCOLS:
75
+ ## CONTEXT BOUNDARIES:
76
+ ## MANDATORY SEQUENCE
77
+ ## 🚨 SYSTEM SUCCESS/FAILURE METRICS:
78
+ ```
79
+
80
+ ### Frontmatter Fields
81
+
82
+ | Field | Required | Description |
83
+ |---|---|---|
84
+ | `name` | Yes | Step identifier |
85
+ | `description` | Yes | One-line purpose |
86
+ | `nextStepFile` | Steps 1–(N-1) | Full path to next step in chain |
87
+ | `outputFile` | If writing output | Path to artifact file (uses `{planning_artifacts}` variable) |
88
+ | `templateFile` | If loading reference | Path to template loaded during execution |
89
+ | `workflowFile` | Final step + cancel paths | Path to `workflow.md` for return-to-menu |
90
+ | `advancedElicitationTask` | Interactive steps | Path to Advanced Elicitation workflow |
91
+ | `partyModeWorkflow` | Interactive steps | Path to Party Mode workflow |
92
+
93
+ ### Step Types
94
+
95
+ | Type | Menu | Use Case | Example |
96
+ |---|---|---|---|
97
+ | **Type 4** (Standard) | A, P, C + custom commands | Interactive steps with user decisions | `step-t-03-score.md` |
98
+ | **Type 5** (Simple) | C only | Non-interactive load/setup steps | `step-t-01-ingest.md` |
99
+ | **Type 10** (Final) | None (auto-return) | Last step — writes output, returns to menu | `step-t-04-update.md` |
100
+
101
+ **Final step convention:** No `nextStepFile`. Has `workflowFile` instead. After completion summary, loads and executes `{workflowFile}` to return to the mode menu.
102
+
103
+ ### Step Chaining
104
+
105
+ Steps form a sequential chain within each mode:
106
+
107
+ ```
108
+ workflow.md → step-[mode]-01 → step-[mode]-02 → ... → step-[mode]-NN → workflow.md
109
+ ```
110
+
111
+ Each step hardcodes the full path to the next step in its menu handling logic (consistent with the `nextStepFile` frontmatter). The final step reloads `workflow.md` via `{workflowFile}`.
112
+
113
+ ---
114
+
115
+ ## Templates
116
+
117
+ Templates are pure markdown reference documents loaded by steps at runtime. They contain no template variables or runtime logic.
118
+
119
+ - **Audience separation:** Create separate templates for different purposes (e.g., one for scoring methodology, one for file format specification)
120
+ - **Installer-managed:** `templates/` is overwritten on update. Users should not customize template files directly — customizations belong in the target project's output folder
121
+ - **Round-trip parseability:** If a template defines a file format, the workflow must be able to reload its own output
122
+
123
+ ---
124
+
125
+ ## Agent Menu Patching
126
+
127
+ The installer adds an `<item>` tag to the target agent's menu:
128
+
129
+ ```xml
130
+ <item cmd="IB or fuzzy match on initiatives-backlog"
131
+ exec="{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md">
132
+ [IB] 📦 Initiatives Backlog (Convoke Enhance)
133
+ </item>
134
+ ```
135
+
136
+ **Conventions:**
137
+ - `cmd`: Short code + fuzzy match phrase (e.g., `"SH or fuzzy match on sprint-health"`)
138
+ - `exec`: Full path using `{project-root}/_bmad/...` form
139
+ - Display label: `[CODE] 📦 [Name] (Convoke Enhance)` — the 📦 prefix and "(Convoke Enhance)" suffix identify enhanced menu items
140
+ - The installer inserts the tag before `</menu>` in the target agent file
141
+
142
+ **Max 2-3 items per agent.** More than that creates menu bloat and dilutes the agent's focus. If you need more capabilities, consider whether they belong as modes within an existing workflow rather than separate workflows.
143
+
144
+ ### Extension Descriptor
145
+
146
+ Create `extensions/bmm-[agent].yaml` to document the menu patch:
147
+
148
+ ```yaml
149
+ # v1: Documentation-only — describes the patch added by the installer
150
+ # v3: Will be read dynamically by agents via <extensions> tag (future BMAD Method proposal)
151
+
152
+ target_agent: bmm/pm
153
+ menu_items:
154
+ - cmd: "IB or fuzzy match on initiatives-backlog"
155
+ exec: "{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md"
156
+ label: "[IB] 📦 Initiatives Backlog (Convoke Enhance)"
157
+ ```
158
+
159
+ ---
160
+
161
+ ## Config Registration
162
+
163
+ Register every workflow in `_bmad/bme/_enhance/config.yaml`:
164
+
165
+ ```yaml
166
+ name: enhance
167
+ version: 1.0.0
168
+ description: "Enhance module — capability upgrades for existing BMAD agents"
169
+ workflows:
170
+ - name: initiatives-backlog
171
+ entry: workflows/initiatives-backlog/workflow.md
172
+ target_agent: bmm/agents/pm.md
173
+ menu_patch_name: "initiatives-backlog"
174
+ ```
175
+
176
+ **Required fields per workflow:**
177
+
178
+ | Field | Description | Example |
179
+ |---|---|---|
180
+ | `name` | Workflow identifier (matches folder name) | `initiatives-backlog` |
181
+ | `entry` | Entry point path relative to `_enhance/` | `workflows/initiatives-backlog/workflow.md` |
182
+ | `target_agent` | Agent file path relative to `_bmad/` | `bmm/agents/pm.md` |
183
+ | `menu_patch_name` | Name attribute for installer detection/dedup | `"initiatives-backlog"` |
184
+
185
+ ### Adding a Second Workflow — Example
186
+
187
+ A "Sprint Health Check" enhancement for the SM agent:
188
+
189
+ ```yaml
190
+ workflows:
191
+ - name: initiatives-backlog
192
+ entry: workflows/initiatives-backlog/workflow.md
193
+ target_agent: bmm/agents/pm.md
194
+ menu_patch_name: "initiatives-backlog"
195
+ - name: sprint-health
196
+ entry: workflows/sprint-health/workflow.md
197
+ target_agent: bmm/agents/sm.md
198
+ menu_patch_name: "sprint-health"
199
+ ```
200
+
201
+ This requires creating:
202
+ - `workflows/sprint-health/workflow.md` + step files + templates
203
+ - `extensions/bmm-sm.yaml` (extension descriptor)
204
+ - Menu item: `[SH] 📦 Sprint Health Check (Convoke Enhance)` in `sm.md`
205
+
206
+ ---
207
+
208
+ ## Verification Integration
209
+
210
+ `verifyInstallation()` runs a 6-point check for the Enhance module (see `scripts/update/lib/validator.js` → `validateEnhanceModule()`):
211
+
212
+ | # | Check | What It Validates |
213
+ |---|---|---|
214
+ | 1 | Directory exists | `_bmad/bme/_enhance/` is present |
215
+ | 2 | Entry point resolves | Each `workflows[].entry` file exists on disk |
216
+ | 3 | Menu patch present | Target agent file contains `<item>` with matching `name` attribute |
217
+ | 4 | Config valid | `config.yaml` has required fields (`name`, `version`, `workflows[]`) |
218
+ | 5 | Filesystem consistency | Workflow directory exists for each registered workflow name |
219
+ | 6 | Skill wrapper exists | `.claude/skills/bmad-enhance-{name}/SKILL.md` is present |
220
+
221
+ If the Enhance directory doesn't exist, all checks pass (Enhance is optional). If it exists, all 6 checks must pass.
222
+
223
+ ### Skill Wrapper
224
+
225
+ The installer generates a skill wrapper at `.claude/skills/bmad-enhance-{workflow-name}/SKILL.md`:
226
+
227
+ ```yaml
228
+ ---
229
+ name: bmad-enhance-{workflow-name}
230
+ description: '[One-line description of what the workflow does]'
231
+ ---
232
+
233
+ IT IS CRITICAL THAT YOU FOLLOW THIS COMMAND: LOAD the FULL {project-root}/_bmad/bme/_enhance/workflows/{workflow-name}/workflow.md, READ its entire contents and follow its directions exactly!
234
+ ```
235
+
236
+ The installer also appends entries to `workflow-manifest.csv` and `skill-manifest.csv` (using `canonicalId` column for dedup).
237
+
238
+ ---
239
+
240
+ ## Quick Verification Checklist
241
+
242
+ Minimum files to create/update for a valid new workflow:
243
+
244
+ - [ ] `_bmad/bme/_enhance/workflows/{name}/workflow.md` — entry point
245
+ - [ ] `_bmad/bme/_enhance/workflows/{name}/steps-*/step-*-*.md` — step files
246
+ - [ ] `_bmad/bme/_enhance/workflows/{name}/templates/*.md` — templates (if needed)
247
+ - [ ] `_bmad/bme/_enhance/config.yaml` — add workflow entry
248
+ - [ ] `_bmad/bme/_enhance/extensions/bmm-{agent}.yaml` — extension descriptor
249
+ - [ ] `.claude/skills/bmad-enhance-{name}/SKILL.md` — skill wrapper (installer creates this)
250
+ - [ ] `workflow-manifest.csv` + `skill-manifest.csv` — manifest entries (installer creates these)
251
+
252
+ Run `verifyInstallation()` after setup — all 6 checks should pass.
@@ -0,0 +1,6 @@
1
+ ---
2
+ name: bmad-enhance-initiatives-backlog
3
+ description: 'Manage RICE initiatives backlog — triage review findings, rescore existing items, or bootstrap new backlogs. Use when the user says triage findings or manage backlog.'
4
+ ---
5
+
6
+ IT IS CRITICAL THAT YOU FOLLOW THIS COMMAND: LOAD the FULL {project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md, READ its entire contents and follow its directions exactly!
@@ -0,0 +1,106 @@
1
+ ---
2
+ name: 'step-c-01-init'
3
+ description: 'Check for existing backlog, warn on overwrite, initialize new backlog using format spec'
4
+ nextStepFile: '{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-02-gather.md'
5
+ outputFile: '{planning_artifacts}/initiatives-backlog.md'
6
+ templateFile: '{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/templates/backlog-format-spec.md'
7
+ workflowFile: '{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/workflow.md'
8
+ ---
9
+
10
+ # Step 1: Initialization & Existing File Guard
11
+
12
+ ## STEP GOAL:
13
+
14
+ Check whether a backlog file already exists at the output location, warn the user if so, and initialize a new backlog session using the backlog format specification.
15
+
16
+ ## MANDATORY EXECUTION RULES (READ FIRST):
17
+
18
+ ### Universal Rules:
19
+ - 🛑 NEVER generate content without user input at overwrite prompt
20
+ - 📖 CRITICAL: Read this complete step file before taking action
21
+ - 🔄 CRITICAL: When loading next step with 'C', read the entire file
22
+ - 📋 YOU ARE AN INITIALIZATION SPECIALIST setting up a new backlog
23
+
24
+ ### Role Reinforcement:
25
+ - ✅ You are an **initialization specialist** — your job is to guard against accidental overwrites and prepare the session
26
+ - ✅ Do not gather initiatives, score items, or write the backlog file — those are later steps
27
+ - ✅ The user must explicitly confirm before any existing file is overwritten
28
+
29
+ ### Step-Specific Rules:
30
+ - 🎯 Focus ONLY on existing file detection, overwrite confirmation, and session setup
31
+ - 🚫 FORBIDDEN to gather initiatives (that is step-c-02's job)
32
+ - 🚫 FORBIDDEN to score items (that is step-c-03's job)
33
+ - 🚫 FORBIDDEN to write the backlog file (that is step-c-04's job)
34
+ - 💬 Approach: check file, warn if needed, confirm readiness, move on
35
+
36
+ ## EXECUTION PROTOCOLS:
37
+ - 🎯 Follow the MANDATORY SEQUENCE exactly
38
+ - 📖 Load `{templateFile}` (backlog-format-spec.md) for structural reference — the format spec defines the file structure that step-c-04 will generate
39
+
40
+ ## CONTEXT BOUNDARIES:
41
+ - Available context: Enhance config (loaded by workflow.md), existing backlog file (if present), backlog format spec template
42
+ - Focus: File existence check and session initialization only
43
+ - Limits: Do NOT gather, score, or write anything
44
+ - Dependencies: workflow.md C dispatch (Create mode selected)
45
+
46
+ ## MANDATORY SEQUENCE
47
+
48
+ **CRITICAL:** Follow this sequence exactly. Do not skip, reorder, or improvise.
49
+
50
+ ### 0. Check for Existing Backlog File
51
+
52
+ Check if a backlog file exists at `{outputFile}`.
53
+
54
+ - **If found:** Display:
55
+
56
+ > **⚠️ Existing backlog detected at `{outputFile}`.**
57
+ >
58
+ > Create mode will generate a **new** backlog file, replacing the existing one. All current backlog data will be lost.
59
+ >
60
+ > **[Y] Yes, overwrite and create new backlog**
61
+ > **[X] Cancel and return to mode selection**
62
+
63
+ **ALWAYS halt and wait for user input.**
64
+
65
+ - IF Y: Continue to step 1.
66
+ - IF X: Display "Cancelling Create mode." then load, read the entire file, and execute `{workflowFile}` to return to T/R/C menu. **Do NOT proceed further.**
67
+ - IF any other input: Display "Please select **Y** or **X**." then redisplay the prompt.
68
+
69
+ - **If NOT found:** Continue to step 1 silently.
70
+
71
+ ### 1. Load Backlog Format Specification
72
+
73
+ Load `{templateFile}` (backlog-format-spec.md) and internalize:
74
+ - **Section hierarchy** — All 7 required H2 sections and their order
75
+ - **Table formats** — Category table (10 columns), Prioritized View (6 columns), Exploration Candidates (4 columns)
76
+ - **Category names** — Existing convention (Documentation & Onboarding, Update & Migration System, etc.)
77
+ - **Item ID format** — Category prefix letter + sequential number
78
+ - **Provenance format** — "Added from Create mode, [date]"
79
+ - **Changelog format** — Table with Date and Change columns, newest first
80
+
81
+ ### 2. Confirm Session Ready
82
+
83
+ Display:
84
+
85
+ > **Create Mode — New Backlog Initialized**
86
+ >
87
+ > A new RICE-scored backlog will be created at `{outputFile}`.
88
+ >
89
+ > **Next:** You'll describe your initiatives one at a time — title, description, and category. Then we'll score them using RICE methodology and generate your prioritized backlog.
90
+
91
+ ### 3. Present MENU OPTIONS
92
+
93
+ Display: "**Select:** [C] Continue to initiative gathering"
94
+
95
+ #### Menu Handling Logic:
96
+ - IF C: Load, read the entire file, and execute `{project-root}/_bmad/bme/_enhance/workflows/initiatives-backlog/steps-c/step-c-02-gather.md`
97
+ - IF any other input: Display "Enter **C** to continue to initiative gathering." then redisplay menu
98
+
99
+ #### EXECUTION RULES:
100
+ - ALWAYS halt and wait for user input after presenting the menu
101
+ - ONLY proceed to next step when user selects 'C'
102
+
103
+ ## 🚨 SYSTEM SUCCESS/FAILURE METRICS:
104
+ ### ✅ SUCCESS: Existing file check performed, overwrite warning shown if needed, format spec loaded, session initialized, proceeding to step-c-02
105
+ ### ❌ SYSTEM FAILURE: Existing file overwritten without warning, format spec not loaded, initiatives gathered prematurely, user not given overwrite choice
106
+ **Master Rule:** Skipping steps is FORBIDDEN.