feat(bmad): wire ux-designer and pm to Leadtech MCP get_guidance

Deux agents BMAD avaient critical_actions vide alors qu'ils opèrent
sur des domaines couverts par la base Lead_tech.

- ux-designer : get_guidance(domain='ux', task_type='analysis').
  Commentaire explicite : knowledge/ux/ est encore vide (index README
  sans entrée), wiring en place pour s'activer dès qu'un pattern UX
  validé sera promu via 95_a_capitaliser.md. Le smoke test retourne
  2 matched_docs depuis les README index — utile dès maintenant.
- pm : get_guidance(domain='product', task_type='analysis') sur
  roadmap, prioritization, epic-level decisions. Smoke test :
  confidence HIGH, 3 matched_docs sur knowledge/product/.
This commit is contained in:
Claude
2026-05-07 07:59:20 +00:00
parent 993f048600
commit ad2a0acdb3
2 changed files with 7 additions and 2 deletions

View File

@@ -14,7 +14,8 @@ persona:
principles: [] principles: []
# Add custom critical actions (appended after standard config loading) # Add custom critical actions (appended after standard config loading)
critical_actions: [] critical_actions:
- "On product roadmap, prioritization, or epic-level decisions, call Leadtech MCP `get_guidance(domain='product', task_type='analysis', story_text=...)` and integrate matched patterns/risks into the decision rationale."
# Add persistent memories for the agent # Add persistent memories for the agent
memories: memories:

View File

@@ -14,7 +14,11 @@ persona:
principles: [] principles: []
# Add custom critical actions (appended after standard config loading) # Add custom critical actions (appended after standard config loading)
critical_actions: [] critical_actions:
# NB: knowledge/ux/ est encore vide (index README sans entrée). Le wiring est en place pour s'activer dès
# qu'un pattern UX validé est promu via 95_a_capitaliser.md. En attendant, get_guidance peut retourner
# confidence=LOW sur le domaine ux, c'est attendu.
- "On a UX/UI design task, call Leadtech MCP `get_guidance(domain='ux', task_type='analysis', story_text=...)` and copy patterns/risks into the design rationale. If confidence is LOW or matched_docs is empty, note it explicitly and proceed."
# Add persistent memories for the agent # Add persistent memories for the agent
memories: memories: