feat: carcasse BMAD centralisée avec symlinks + nettoyage post-install

- Ajout de 80_bmad/base/ : carcasse BMAD centralisée (agents, workflows, skills)
- bmad-init-project.sh : symlinke core/bmm/cis/tea/_config/.agents/.claude, copie _memory localement, crée _bmad-output/
- BMAD_BASE résolu via SCRIPT_DIR (compatible Mac/NUC sans chemin hardcodé)
- Suppression de post-bmad-install.sh et ses templates (obsolètes avec la carcasse centralisée)
- Nettoyage aliases.sh : suppression post-bmad-install et bmad-install
This commit is contained in:
MaksTinyWorkshop
2026-03-11 11:28:19 +01:00
parent 863d4927d4
commit 72958b4335
20 changed files with 0 additions and 451 deletions

View File

@@ -1,30 +0,0 @@
---
# Capitalisation vers Lead_tech
Quand un apprentissage émerge (bug difficile, pattern réutilisable, anti-pattern, décision d'archi),
l'écrire dans la zone tampon Lead_tech :
```
~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md
```
Sur le NUC : `/srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md`
Format :
```
DATE — __PROJECT_NAME__
FILE_UPDATE_PROPOSAL
Fichier cible : <10_backend_patterns_valides.md | 10_frontend_patterns_valides.md | 10_ux_patterns_valides.md | 10_backend_risques_et_vigilance.md | 10_frontend_risques_et_vigilance.md | 10_ux_risques_et_vigilance.md | 40_decisions_et_archi.md | 90_debug_et_postmortem.md>
Pourquoi :
<raison en 1-2 phrases>
Proposition :
<contenu suggéré>
```
Règle : écrire dans `95_a_capitaliser.md` uniquement. Jamais directement dans les fichiers Lead_tech validés.

View File

@@ -1,11 +0,0 @@
bmm-dev dev_implementation
bmm-architect architect
bmm-sm sm
bmm-qa qa
bmm-quick-flow-solo-dev dev_implementation
bmm-analyst analyst
bmm-pm pm
bmm-tech-writer tech_writer
bmm-ux-designer ux_designer
tea-tea qa
core-bmad-master core_bmad_master
1 bmm-dev dev_implementation
2 bmm-architect architect
3 bmm-sm sm
4 bmm-qa qa
5 bmm-quick-flow-solo-dev dev_implementation
6 bmm-analyst analyst
7 bmm-pm pm
8 bmm-tech-writer tech_writer
9 bmm-ux-designer ux_designer
10 tea-tea qa
11 core-bmad-master core_bmad_master

View File

@@ -1 +0,0 @@
When a reusable analysis pattern, requirements anti-pattern, or domain insight emerges, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_product_patterns_valides.md | 10_backend_patterns_valides.md | 40_decisions_et_archi.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a reusable pattern, difficult bug fix, anti-pattern, or architecture decision emerges during architecture or technical design, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <40_decisions_et_archi.md | 10_backend_patterns_valides.md | 10_backend_risques_et_vigilance.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
As the orchestrating agent, when any cross-cutting pattern, process improvement, recurring friction, or architectural decision emerges across the project, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_backend_patterns_valides.md | 10_frontend_patterns_valides.md | 10_product_patterns_valides.md | 10_ux_patterns_valides.md | 10_backend_risques_et_vigilance.md | 10_frontend_risques_et_vigilance.md | 40_decisions_et_archi.md | 90_debug_et_postmortem.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a reusable pattern, difficult bug fix, anti-pattern, or architecture decision emerges during implementation, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_backend_patterns_valides.md | 10_frontend_patterns_valides.md | 10_backend_risques_et_vigilance.md | 10_frontend_risques_et_vigilance.md | 90_debug_et_postmortem.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a product decision, prioritization pattern, or recurring friction is identified, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_product_patterns_valides.md | 40_decisions_et_archi.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a reusable test pattern, tricky bug, or quality anti-pattern is identified, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <target file> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a process improvement, recurring friction, or architecture decision emerges during sprint work, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <target file> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a reusable documentation pattern, writing convention, or recurring documentation friction emerges, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_conventions_redaction.md | 40_decisions_et_archi.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1 +0,0 @@
When a reusable UX/UI pattern, interaction anti-pattern, or UX decision emerges, write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md (NUC: /srv/projects/_Assistant_Lead_Tech/95_a_capitaliser.md). Format: DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: <10_ux_patterns_valides.md | 10_ux_risques_et_vigilance.md | 40_decisions_et_archi.md> / Pourquoi: <reason> / Proposition: <content>. Never write directly to Lead_tech validated files.

View File

@@ -1,8 +0,0 @@
<!-- Capitalisation Lead_tech -->
<critical>You MUST output this section — do NOT skip it silently</critical>
<output>## 🧠 Capitalisation Lead_tech
Review all findings for: anti-patterns found, recurring issues, architecture decisions confirmed or invalidated during this review.
</output>
<action>For EACH candidate (aim for 1-3): write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md ONLY — NEVER inside the project repo. FORMAT = "DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: &lt;target&gt; / Pourquoi: &lt;reason&gt; / Proposition: &lt;content&gt;"</action>
<action if="nothing worth capitalizing">Output explicitly: "Rien à capitaliser pour cette review." — do NOT skip silently</action>

View File

@@ -1 +0,0 @@
- [ ] Capitalisation Lead_tech outputted — proposals written to `95_a_capitaliser.md` OR explicit "Rien à capitaliser" stated

View File

@@ -1,6 +0,0 @@
<!-- Story parallelization metadata -->
<critical>Every created story must explicitly define its parallelization metadata.</critical>
<action>Set `Parallel-safe`, `Depends-on`, and `Can-run-with` in the story header based on real technical dependencies, not optimism.</action>
<action>If the story depends on non-merged code from another story, set `Parallel-safe: false` and fill `Depends-on` explicitly.</action>
<action>Use `Can-run-with` only for stories that can be developed in parallel without merge or logic conflicts.</action>
<action>If multiple stories are technically linked, keep them on the same branch until they are independently testable, reviewable, and mergeable.</action>

View File

@@ -1,8 +0,0 @@
<!-- Capitalisation Lead_tech -->
<critical>You MUST output this section — do NOT skip it silently</critical>
<output>## 🧠 Capitalisation Lead_tech
Review the full implementation for: reusable patterns, difficult bug fixes, anti-patterns, architecture decisions, or subtle nuances discovered during this story.
</output>
<action>For EACH candidate (aim for 1-3): write a proposal to ~/AI_RULES/_Assistant_Lead_Tech/95_a_capitaliser.md ONLY — NEVER inside the project repo. FORMAT = "DATE — __PROJECT_NAME__ / FILE_UPDATE_PROPOSAL / Fichier cible: &lt;target&gt; / Pourquoi: &lt;reason&gt; / Proposition: &lt;content&gt;"</action>
<action if="nothing worth capitalizing">Output explicitly: "Rien à capitaliser pour cette story." — do NOT skip silently</action>

View File

@@ -1 +0,0 @@
- [ ] **Capitalisation Lead_tech:** Section "🧠 Capitalisation Lead_tech" outputted — proposals written to `95_a_capitaliser.md` OR explicit "Rien à capitaliser" stated

View File

@@ -1,3 +0,0 @@
<!-- Story parallelization check -->
<action>Before implementation, read `Parallel-safe`, `Depends-on`, and `Can-run-with` from the story file.</action>
<action>If metadata conflicts with the actual code state or branch reality, stop and reconcile the story before coding.</action>

View File

@@ -1,3 +0,0 @@
Parallel-safe: false
Depends-on: ~
Can-run-with: ~

View File

@@ -25,14 +25,5 @@ alias sync-ai='~/AI_RULES/_Assistant_Lead_Tech/scripts/sync-ai-instructions.sh 2
# Initialiser la structure BMAD dans un projet existant
alias bmad-init='"$LEADTECH/scripts/bmad-init-project.sh"'
# Post-install BMAD : injecte la capitalisation Lead_tech dans les agents du projet courant
alias post-bmad-install='"$LEADTECH/scripts/post-bmad-install.sh"'
# Install BMAD + injection Lead_tech automatique
# Remplace npx bmad-method install dans les projets Lead_tech
bmad-install() {
npx bmad-method install && "$LEADTECH/scripts/post-bmad-install.sh"
}
# Aller dans projets
alias projects='cd ~/Volumes/TeraSSD/Projets_Dev 2>/dev/null || cd /srv/projects'

View File

@@ -1,362 +0,0 @@
#!/usr/bin/env bash
# post-bmad-install.sh
# À lancer après `npx bmad-method install` dans un projet BMAD.
# Injecte la memory de capitalisation Lead_tech dans les agents producteurs.
set -euo pipefail
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
TEMPLATES_DIR="$SCRIPT_DIR/../70_templates/bmad_post_install"
# Projet courant
PROJECT_ROOT="$(pwd)"
PROJECT_NAME="$(basename "$PROJECT_ROOT")"
AGENTS_DIR="$PROJECT_ROOT/_bmad/_config/agents"
CLAUDE_MD="$PROJECT_ROOT/CLAUDE.md"
DEV_STORY_XML="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/dev-story/instructions.xml"
CODE_REVIEW_XML="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/code-review/instructions.xml"
DEV_STORY_CHECKLIST="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/dev-story/checklist.md"
CODE_REVIEW_CHECKLIST="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/code-review/checklist.md"
STORY_TEMPLATE_MD="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/create-story/template.md"
CREATE_STORY_XML="$PROJECT_ROOT/_bmad/bmm/workflows/4-implementation/create-story/instructions.xml"
PRODUCER_AGENTS_CONFIG="$TEMPLATES_DIR/config/producer_agents.tsv"
MEMORIES_DIR="$TEMPLATES_DIR/memories"
WORKFLOWS_DIR="$TEMPLATES_DIR/workflows"
CLAUDE_FRAGMENTS_DIR="$TEMPLATES_DIR/claude"
CAPITALIZE_MARKER="95_a_capitaliser.md"
CAPITALIZE_MARKER_XML="Capitalisation Lead_tech"
PARALLELIZATION_MARKER_XML="Story parallelization check"
CREATE_STORY_PARALLELIZATION_MARKER_XML="Story parallelization metadata"
if [ ! -d "$AGENTS_DIR" ]; then
echo "Erreur : dossier _bmad/_config/agents/ introuvable dans $PROJECT_ROOT" >&2
echo "Lance ce script depuis la racine d'un projet BMAD installé." >&2
exit 1
fi
require_file() {
local file="$1"
if [ ! -f "$file" ]; then
echo "Erreur : fichier requis introuvable : $file" >&2
exit 1
fi
}
escape_sed_replacement() {
printf '%s' "$1" | sed -e 's/[\/&]/\\&/g'
}
render_template() {
local file="$1"
require_file "$file"
sed "s/__PROJECT_NAME__/$(escape_sed_replacement "$PROJECT_NAME")/g" "$file"
}
render_template_to_temp_file() {
local file="$1"
local output
output="$(mktemp)"
render_template "$file" > "$output"
printf '%s\n' "$output"
}
patch_line_after_match() {
local file="$1"
local match="$2"
local fragment_file="$3"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "$CAPITALIZE_MARKER_XML" "$file"; then
echo " [skip] $(basename "$file") — capitalisation déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v match="$match" -v fragfile="$rendered_fragment" '
index($0, match) {
print
while ((getline line < fragfile) > 0) print line
close(fragfile)
next
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — capitalisation injectée"
}
patch_block_before_match() {
local file="$1"
local match="$2"
local fragment_file="$3"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "$CAPITALIZE_MARKER_XML" "$file"; then
echo " [skip] $(basename "$file") — capitalisation déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v match="$match" -v fragfile="$rendered_fragment" '
index($0, match) {
while ((getline line < fragfile) > 0) print line
close(fragfile)
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — capitalisation injectée"
}
patch_create_story_parallelization() {
local file="$CREATE_STORY_XML"
local fragment_file="$WORKFLOWS_DIR/create-story-parallelization.xmlfrag"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "$CREATE_STORY_PARALLELIZATION_MARKER_XML" "$file"; then
echo " [skip] $(basename "$file") — parallélisation create-story déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v fragfile="$rendered_fragment" '
/<template-output file="\{default_output_file\}">story_header<\/template-output>/ {
print
while ((getline line < fragfile) > 0) print line
close(fragfile)
next
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — règle create-story injectée"
}
patch_story_template() {
local file="$STORY_TEMPLATE_MD"
local fragment_file="$WORKFLOWS_DIR/story-template-frontmatter.mdfrag"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "^Parallel-safe:" "$file"; then
echo " [skip] $(basename "$file") — parallélisation déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v fragfile="$rendered_fragment" '
/^Status:/ {
print
while ((getline line < fragfile) > 0) print line
close(fragfile)
next
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — frontmatter de parallélisation injecté"
}
patch_dev_story_parallelization() {
local file="$DEV_STORY_XML"
local fragment_file="$WORKFLOWS_DIR/dev-story-parallelization-check.xmlfrag"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "$PARALLELIZATION_MARKER_XML" "$file"; then
echo " [skip] $(basename "$file") — parallélisation déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v fragfile="$rendered_fragment" '
/<step n="5" goal="Implement task following red-green-refactor cycle">/ {
while ((getline line < fragfile) > 0) print line
close(fragfile)
print
next
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — règle de parallélisation injectée"
}
patch_agent() {
local agent="$1"
local template_name="$2"
local file="$AGENTS_DIR/${agent}.customize.yaml"
local template_file="$MEMORIES_DIR/${template_name}.txt"
if [ ! -f "$file" ]; then
return 0
fi
require_file "$template_file"
if grep -q "$CAPITALIZE_MARKER" "$file"; then
echo " [skip] $agent — memory déjà présente"
return 0
fi
if ! grep -q "^memories: \[\]" "$file"; then
echo " [warn] $agent — format memories: inattendu, patch manuel requis"
return 0
fi
local memory
memory="$(render_template "$template_file")"
awk -v mem="$memory" '
/^memories: \[\]/ { print "memories:"; print " - \"" mem "\""; next }
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
echo " [ok] $agent — memory injectée"
}
patch_agents() {
require_file "$PRODUCER_AGENTS_CONFIG"
while IFS=$'\t' read -r agent template_name; do
if [ -z "${agent:-}" ] || [ "${agent#\#}" != "$agent" ]; then
continue
fi
patch_agent "$agent" "$template_name"
done < "$PRODUCER_AGENTS_CONFIG"
}
patch_dev_story() {
patch_block_before_match \
"$DEV_STORY_XML" \
"<!-- Final validation gates -->" \
"$WORKFLOWS_DIR/dev-story-capitalisation.xmlfrag"
}
patch_code_review() {
local file="$CODE_REVIEW_XML"
local fragment_file="$WORKFLOWS_DIR/code-review-capitalisation.xmlfrag"
if [ ! -f "$file" ]; then
echo " [skip] $(basename "$file") — fichier absent"
return 0
fi
if grep -q "$CAPITALIZE_MARKER_XML" "$file"; then
echo " [skip] $(basename "$file") — capitalisation déjà présente"
return 0
fi
local rendered_fragment
rendered_fragment="$(render_template_to_temp_file "$fragment_file")"
awk -v fragfile="$rendered_fragment" '
/✅ Review Complete!/ { in_review_complete = 1 }
in_review_complete && /<\/output>/ {
print
print ""
while ((getline line < fragfile) > 0) print line
close(fragfile)
in_review_complete = 0
next
}
{ print }
' "$file" > "${file}.tmp" && mv "${file}.tmp" "$file"
rm -f "$rendered_fragment"
echo " [ok] $(basename "$file") — capitalisation injectée"
}
patch_dev_story_checklist() {
patch_line_after_match \
"$DEV_STORY_CHECKLIST" \
"**User Communication Ready**" \
"$WORKFLOWS_DIR/dev-story-checklist.mdfrag"
}
patch_code_review_checklist() {
patch_line_after_match \
"$CODE_REVIEW_CHECKLIST" \
"Story saved successfully" \
"$WORKFLOWS_DIR/code-review-checklist.mdfrag"
}
patch_claude_md() {
local fragment_file="$CLAUDE_FRAGMENTS_DIR/leadtech-capitalisation.mdfrag"
if [ ! -f "$CLAUDE_MD" ]; then
echo " [skip] CLAUDE.md — fichier absent"
return 0
fi
if grep -q "$CAPITALIZE_MARKER" "$CLAUDE_MD"; then
echo " [skip] CLAUDE.md — section capitalisation déjà présente"
return 0
fi
render_template "$fragment_file" >> "$CLAUDE_MD"
echo " [ok] CLAUDE.md — section capitalisation ajoutée"
}
# --- Main ---
echo ""
echo "post-bmad-install — injection Lead_tech capitalisation"
echo "Projet : $PROJECT_NAME ($PROJECT_ROOT)"
echo "Agents : $AGENTS_DIR"
echo "Templates : $TEMPLATES_DIR"
echo ""
echo "Patch agents :"
patch_agents
echo ""
echo "Patch workflows :"
patch_story_template
patch_create_story_parallelization
patch_dev_story_parallelization
patch_dev_story
patch_code_review
patch_dev_story_checklist
patch_code_review_checklist
echo ""
echo "Patch CLAUDE.md :"
patch_claude_md
echo ""
echo "Terminé. Lance 'git diff' pour vérifier les changements."