bmc-dashboard-audit-runner

Runs the full BMC Dashboard audit at depth. Ensures stack is up, executes run_audit.sh, probes all API endpoints, and writes report + handoff to .cursor/bmc-audit/ for the Debug Reviewer. Use when use

SKILL.md

GitHub

BMC Dashboard Audit Runner Skill

Before working: Read docs/team/knowledge/AuditDebug.md if it exists.

Coordinates the Audit Runner Agent to execute a thorough BMC Dashboard audit.

When to Use

Trigger Action
"run audit a fondo" Execute full audit
"bmc audit completo" Same
"full dashboard audit" Same
Before Debug Reviewer Audit Runner must complete first

Execution Checklist

  1. Start stack if API (3001) not responding
  2. Run bash .cursor/skills/super-agente-bmc-dashboard/scripts/run_audit.sh --output=.cursor/bmc-audit/latest-report.md
  3. Probe all endpoints (health, cotizaciones, proximas-entregas, pagos-pendientes, kpi-financiero, audit, calc/catalogo)
  4. Write handoff.json to .cursor/bmc-audit/
  5. Emit "AUDIT RUNNER COMPLETE. Invoke bmc-dashboard-debug-reviewer."
  6. Update docs/team/PROJECT-STATE.md (Cambios recientes) con resumen del audit.

Output Paths

File Purpose
.cursor/bmc-audit/latest-report.md Full report from run_audit.sh
.cursor/bmc-audit/handoff.json Metadata for Debug Reviewer
.cursor/bmc-audit/endpoint-probe.json Optional: raw endpoint responses

One-Shot Script (Audit + Debug)

To run both Audit Runner and Debug Reviewer sequentially:

bash .cursor/skills/bmc-dashboard-audit-runner/scripts/run_audit_then_debug.sh

Requires API running on 3001. Output: .cursor/bmc-audit/

Reference