fix(154): MAESTRO layer detection in /tachi.infographic checks wrong file and pattern

The "all" template expansion greps for "MAESTRO Layer" in the richest data
source file (compensating-controls.md or risk-scores.md), but MAESTRO layer
assignments live in threats.md. The grep pattern also matches only column
headers ("| MAESTRO Layer |"), not actual layer values ("L3 — Agent Framework").

Fix: check threats.md specifically (regardless of richest tier), and grep
for actual layer values with `L[1-7] [—–-]` which matches data rows.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: davidmatousek

.claude/commands/tachi.infographic.md

@@ -161,7 +161,7 @@ Single-command entry point for tachi threat infographic generation — the visua
        --output /tmp/infographic-{template_name}.json
 
    Template expansion:
-   - "all" expands to: baseball-card, system-architecture, risk-funnel, executive-architecture. Then check if data source contains MAESTRO layer data (grep for "MAESTRO Layer" or "maestro_layer" in the data source file) — if present, also append maestro-stack and maestro-heatmap (up to 6 runs total).
+   - "all" expands to: baseball-card, system-architecture, risk-funnel, executive-architecture. Then check if threats.md contains MAESTRO layer data by grepping for actual layer values (NOT column headers): `grep -P 'L[1-7] [—–-]' {data_source_dir}/threats.md` — match any of em-dash, en-dash, or hyphen after L1-L7. If the pattern matches at least one data row (not just a header), also append maestro-stack and maestro-heatmap (up to 6 runs total). IMPORTANT: always check threats.md for MAESTRO detection regardless of which file is the richest data source — MAESTRO layer assignments are in the threats.md component table and findings tables, not in risk-scores.md or compensating-controls.md.
    - "maestro" expands to: maestro-stack, maestro-heatmap (2 runs, sequential)
    - "exec" resolves to: executive-architecture (alias, 1 run)
    - Individual templates (baseball-card, system-architecture, risk-funnel, maestro-stack, maestro-heatmap, executive-architecture): 1 run