| Métrica | this 7d | prior 7d | Δ |
|---|---|---|---|
| Activas | 43 | 51 | -15.7% |
| Nuevas | 25 | 36 | -30.6% |
| Link sent | 0 | 0 | — |
| Calls booked | 0 | 0 | — |
| Booked/Link % | — | — | — |
| Booked/Activas % | 0.0% | 0.0% | +0.0pp |
| Booked/Nuevas % | 0.0% | 0.0% | +0.0pp |
| Revenue variable | $1.30 | $3.01 | -56.8% |
| Período | Activas | Nuevas | Link sent | Calls booked | Booked/Link | Booked/Activas | Booked/Nuevas |
|---|---|---|---|---|---|---|---|
| Semana actual (7d) | 43 | 25 | 0 | 0 | — | 0.0% | 0.0% |
| Semana anterior (7d) | 51 | 36 | 0 | 0 | — | 0.0% | 0.0% |
| Día | Activas | Nuevas | Link sent | Calls booked | Booked/Link | Booked/Activas | Booked/Nuevas |
|---|---|---|---|---|---|---|---|
| 06-02 | 4 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 06-01 | 4 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 05-31 | 6 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 05-30 | 14 | 6 | 0 | 0 | — | 0.0% | 0.0% |
| 05-29 | 7 | 4 | 0 | 0 | — | 0.0% | 0.0% |
| 05-28 | 3 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 05-27 | 5 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 05-26 | 10 | 8 | 0 | 0 | — | 0.0% | 0.0% |
| 05-25 | 5 | 4 | 0 | 0 | — | 0.0% | 0.0% |
| 05-24 | 9 | 7 | 0 | 0 | — | 0.0% | 0.0% |
| 05-23 | 6 | 3 | 0 | 0 | — | 0.0% | 0.0% |
| 05-22 | 6 | 2 | 0 | 0 | — | 0.0% | 0.0% |
| 05-21 | 7 | 6 | 0 | 0 | — | 0.0% | 0.0% |
| 05-20 | 8 | 6 | 0 | 0 | — | 0.0% | 0.0% |
| BP | Sev | Estado | Detalle |
|---|---|---|---|
| BP-1 | HIGH | N/A | Template compliance meta.md declares Template: TBD and no matching template file was provided. BP-1 cannot be evaluated without a declared template to verify against. |
| BP-2 | MEDIUM | PASS | Example coverage and cap (7 base scenarios + conditional, ≤20 total) examples.md contains 13 examples covering all 7 base scenarios (happy path, resource/link delivery, objection handling, graceful disqualification, emotional/sensitive lead, re-engagement, off-topic question) plus keyword-trigger and price-objection conditionals. Total count (13) is well under the 20-example cap. |
| BP-3 | HIGH | PASS | Examples align with the flow All 13 examples in examples.md follow the flow defined in system.md (Apertura → Perfilado → Diagnóstico → Posicionar → Propuesta + Formulario → Post-formulario) and reach terminal states (formulario sent, NO_RESPONSE, or resource shared with closure). |
| BP-4 | CRITICAL | PASS | No internal contradictions in system.md system.md is internally consistent. The flow steps, voice rules, calibration guidance, and NO_RESPONSE conditions do not contradict each other. Stated exceptions (e.g., 'NO_RESPONSE post-formulario') are explicit carve-outs, not contradictions. |
| BP-5 | CRITICAL | PASS | No cross-file contradictions All SDK files align with system.md. The program.json note flagging 'DRIFT a resolver con Rita' (method name mismatch: O.E.C.V.™ vs. Arquitectura de Reputación©) is an internal developer annotation, not a behavioral rule or data point that fires during conversation. Per the dismissed false positives list, this metadata annotation does not produce conflicting agent outputs under any shared precondition. |
| BP-6 | MEDIUM | PASS | Voice consistency across examples All 13 examples use consistent Rioplatense voice (voseo, short sentences, warm tone, validation rotation). No jarring tonal shifts; all examples sound like the same agent. |
| BP-7 | CRITICAL | PASS | Unguarded actions (PAT-004) All risky actions (sending formulario, NO_RESPONSE, sharing resources) are guarded by explicit conditions in system.md. For example, formulario is sent 'SOLO cuando el lead aceptó la llamada explícitamente' and post-formulario NO_RESPONSE is stated clearly. Examples demonstrate guards firing correctly (e.g., Example 0 shows lead saying 'Sí' before formulario is sent). |
| BP-8 | HIGH | PASS | No stale version language in examples All example CTAs and closing language match the current flow definition in system.md. The formulario text, validation phrases, and NO_RESPONSE patterns are consistent across examples and system.md. |
| BP-9 | MEDIUM | PASS | No duplicated instructions Each behavioral instruction has a single canonical home. For example, the 'one question per message' rule appears in Principios (canonical) and is demonstrated in examples, not restated. No instruction is duplicated across sections. |
| BP-10 | MEDIUM | PASS | Rule-vs-principle balance Behavioral guidance (tone, flow, conversational style) is expressed via Voz section + Principios + worked examples. Hard rules remain for factual constraints (price never in chat, formulario links, NO_RESPONSE conditions). The balance is appropriate. |
| Check | Sec | Estado | Notas |
|---|---|---|---|
| 6a | 1.8 | FAIL | forbidden tokens found: ['PARAR', 'STOP'] |
| 6c | 3.x | PASS | 118 lines |
| 6d | 1.2 | PASS | 13 examples |
| 6e | 1.3 | PASS | 7 principles |
| 6g | PAT-006 | WARN | non-STEP-0 keyword handling |
| 6h | Anti-pat | WARN | 1 'burbuja' mentions |
| 6i | Anti-pat | PASS | no HTML comments |
| 6j | Anti-pat | FAIL | 12 em-dashes across 3 file(s); first: system.md:15 |
| 6k | PAT-006 | PASS | no tool syntax leak |
| 6l | Resources | PASS | no placeholder URLs |
| 6m | Hygiene | PASS | no audit metadata in prompt files |
| 6n | PAT-008 | WARN | 2 URL(s) appear with different tracking params or formats across SDK files |
| 6q | PAT-001 | PASS | no instruction duplication into examples.md, keywords.json, program.json, resources.json |
| 6r | Keyword-drift | FAIL | 6 in keywords.json + recent digest, never named in system.md: ['MANDAMIENTOS', 'FERRARI', 'VIDEO', 'SISTEMA', 'SECRETO', 'B-ROLL']; (2 stale drift entr(ies) ignored — no recent digest mention) |
| 6s | Hygiene | PASS | no expired dates in live SDK |
| 6t | Coverage | WARN | 7/7 principle(s) without example: #1 Una respuesta tuya por, #2 No deduzcas lo que, #3 Una pregunta por mensaje, #4 No precios en chat, #5 No tenes que agendár; 1/6 flow step(s) without example: #5 Propuesta + Formulario |
| Fecha | Tema | Tags |
|---|---|---|
| 2026-06-02 | manychat ninjo integration clarification | architecture, integration, manychat, ninjo |
| 2026-06-02 | agent conversation quality issues | agent-behavior, bug, conversation-flow, scheduling |
| 2026-06-02 | bot corta repite no agenda | bug, repetition, broken-sequence, no-booking, feedback, reports |
| 2026-06-02 | weekly reporting customization | reporting, analytics, metrics |
| 2026-06-01 | weekly metrics keyword performance visin hypothesis | metrics, keyword-testing, funnel, content-strategy |
| 2026-05-31 | bot freezing conversation drop | bug, blocking, agent-behavior |
| 2026-05-31 | welcome flow qualification bifurcation | workflow, qualification, prompt-update, messaging-flow |
| 2026-05-29 | benjamin sonnecheck in message | manual-message, low-priority |
| 2026-05-28 | ninjo agent v2 deployment founder voice | agent-deployment, product-change, voice-identity, high-priority |
| 2026-05-28 | ninjo agent v2 deployment and feedback requests | agent-update, deployment, feedback-required, high-priority |
| 2026-05-28 | agent testing and monitoring phase | testing, qa, monitoring, feedback |
| 2026-05-27 | ninjo database state query error | technical-issue, database, blocker |
| 2026-05-26 | ninjo agent improvements status check | ninjo-agent, improvements, status-update |
| 2026-05-26 | manyhat to ninjo flow migration tracking | migration, workflow, deadline |
| 2026-05-25 | ninjo agent booking confirmation policy | agent-behavior, booking-flow, policy |
| 2026-05-25 | weekly funnel report may 18 24 | metrics, performance, keywords, content-strategy |
| 2026-05-25 | weekly funnel report test keyword performance | analytics, keyword-performance, conversion |
| 2026-05-25 | conversation quality concerns agent inference | quality-assurance, agent-behavior, prompt-refinement |
| 2026-05-23 | cortex strategy recap migration plan | strategy, manychat-migration, product-positioning |
| 2026-05-23 | ninjo agent pause on rita input | agent-behavior, configuration, decision |
Sin pendientes en el checklist ✅
| Señal | Valor | Evidencia (cita del cliente) |
|---|---|---|
| Enojado con la plataforma | sí | “Me frustra un poco porque no estamos logrando ni agendar ni tener unas conversaciones buenas.” |
| Es queja | sí | “Me frustra un poco porque no estamos logrando ni agendar ni tener unas conversaciones buenas.” |
| Queja constructiva (accionable) | sí | “Me frustra un poco porque no estamos logrando ni agendar ni tener unas conversaciones buenas.” |
| Señal | Cita textual | Fecha |
|---|---|---|
| Frustración explícita con resultados del agente: no agenda y conversaciones de baja calidad | “Me frustra un poco porque no estamos logrando ni agendar ni tener unas conversaciones buenas.” | 2026-06-02 |
| Reporte de bug crítico: agente repite la misma pregunta ignorando respuestas del lead y rompe la secuencia | “Tenés 30 minutos esta semana para una llamada con mi equipo?” | 2026-06-02 |
| Involucramiento activo: Rita escala con captura de conversación específica (019e07fc/Cori), señal de que sigue apostando a la plataforma | “None” | 2026-06-02 |
| Intervención manual recurrente: Rita retomaba mensajes que el bot dejó colgados, carga operativa directa sobre el cliente | “None” | 2026-06-02 |
| Participación estratégica: Rita propone bifurcación explícita en el flujo de bienvenida con copy propio, señal de compromiso con el producto | “Creo que en esta etapa te va a aportar mucho trabajar primero sobre crecimiento...” | 2026-05-31 |
| Reporte de bug de congelamiento del agente | “el bot queda tildado, no sigue las conversaciones” | 2026-05-31 |
Sin cambios pendientes en el checklist ✅