| Métrica | this 7d | prior 7d | Δ |
|---|---|---|---|
| Activas | 7 | 1 | +600.0% |
| Nuevas | 0 | 0 | — |
| Link sent | 0 | 0 | — |
| Calls booked | 0 | 0 | — |
| Booked/Link % | — | — | — |
| Booked/Activas % | 0.0% | 0.0% | +0.0pp |
| Booked/Nuevas % | — | — | — |
| Revenue variable | $0.00 | $0.00 | — |
| Período | Activas | Nuevas | Link sent | Calls booked | Booked/Link | Booked/Activas | Booked/Nuevas |
|---|---|---|---|---|---|---|---|
| Semana actual (7d) | 7 | 0 | 0 | 0 | — | 0.0% | — |
| Semana anterior (7d) | 1 | 0 | 0 | 0 | — | 0.0% | — |
| Día | Activas | Nuevas | Link sent | Calls booked | Booked/Link | Booked/Activas | Booked/Nuevas |
|---|---|---|---|---|---|---|---|
| 07-30 | 2 | 0 | 0 | 0 | — | 0.0% | — |
| 07-29 | 2 | 0 | 0 | 0 | — | 0.0% | — |
| 07-28 | 2 | 0 | 0 | 0 | — | 0.0% | — |
| 07-27 | 0 | 0 | 0 | 0 | — | — | — |
| 07-26 | 0 | 0 | 0 | 0 | — | — | — |
| 07-25 | 0 | 0 | 0 | 0 | — | — | — |
| 07-24 | 1 | 0 | 0 | 0 | — | 0.0% | — |
| 07-23 | 0 | 0 | 0 | 0 | — | — | — |
| 07-22 | 0 | 0 | 0 | 0 | — | — | — |
| 07-21 | 0 | 0 | 0 | 0 | — | — | — |
| 07-20 | 0 | 0 | 0 | 0 | — | — | — |
| 07-19 | 0 | 0 | 0 | 0 | — | — | — |
| 07-18 | 0 | 0 | 0 | 0 | — | — | — |
| 07-17 | 1 | 0 | 0 | 0 | — | 0.0% | — |
⏳ Sin re-auditar hoy (sin cambios de prompt) — mostrando la última auditoría disponible. Estado vigente hasta nueva corrida.
| BP | Sev | Estado | Detalle |
|---|---|---|---|
| BP-1 | HIGH | PASS | Template compliance Agent declares `conversational` template. system.md contains all required Tier 2 sections (Identity, Voice, Flow, Calibration, Principles, Tools) plus knowledge_base.md and examples.md providing Tier 3 reference material. Structure matches the template specification. |
| BP-2 | MEDIUM | PASS | Example coverage and cap (7 base scenarios + conditional, ≤20 total) examples.md contains 14 examples covering all 7 base scenarios (happy path, resource delivery, objection handling, disqualification, emotional/sensitive, re-engagement, off-topic) plus keyword-trigger example (APPROACH keyword in Example 3). Total count is 14, within the ≤20 cap. evidence (1)
|
| BP-3 | HIGH | PASS | Examples align with the flow All 14 examples follow the flow steps defined in system.md (Open → Discover → Dig → Connect → Resource → Bridge → Close) and reach terminal states (NO_RESPONSE, application submitted, or link sent with conversation ended). evidence (1)
|
| BP-4 | CRITICAL | PASS | No internal contradictions in system.md system.md contains one explicit exception to the consent rule: 'Always get consent before sending any link or resource. Ask first, send only after they say yes. Exception: hot leads who explicitly ask for the link.' This exception is clearly documented and does not contradict the default rule — they apply under different preconditions (hot lead vs. standard lead). evidence (1)
|
| BP-5 | CRITICAL | PASS | No cross-file contradictions keywords.json declares `send_without_permission: true` for 'talk' and 'places' keywords. system.md explicitly documents this exception in the Flow section: 'If a keyword was sent (APPROACH, etc), adjust the resource to match the keyword topic.' The exception is stated in system.md, and keywords.json implements it under a different precondition (keyword-triggered immediate delivery vs. standard DM flow). Both rules coexist without firing simultaneously under the same precondition. evidence (2)
|
| BP-6 | MEDIUM | PASS | Voice consistency across examples All 14 examples use consistent voice: lowercase, no periods, casual buddy-like tone, max 30 words per message, no emojis, no markdown. Tone is uniform across happy path, objections, disqualifications, and edge cases. evidence (1)
|
| BP-7 | CRITICAL | PASS | Unguarded actions (PAT-004) All risky actions (sending links, qualifying, closing) are guarded. Consent rule is stated: 'Always get consent before sending any link or resource' with explicit exception for hot leads. Prequalification gate is stated: 'Minimum: learn where theyre from AND whats going on with their dating life before offering any resource.' Examples demonstrate guards firing correctly (Example 1 shows consent before link; Example 2 shows hot lead exception; Example 6 shows price objection handling). evidence (2)
|
| BP-8 | HIGH | PASS | No stale version language in examples All examples use current language from system.md. CTAs match: 'fill it out and let me know' / 'fill it out and well get you on the calendar' align with system.md's voice examples. NO_RESPONSE closers match the stated rule. No outdated phrasing detected. evidence (2)
|
| BP-9 | MEDIUM | PASS | No duplicated instructions Each behavioral instruction has a single canonical home. Consent rule stated once in Principles. Prequalification rule stated once in Flow. Price handling stated once in Principles. Objection scripts in objections.md do not duplicate system.md rules — they provide specific phrasings for handling objections, not restating the rules themselves. |
| BP-10 | MEDIUM | PASS | Rule-vs-principle balance Behavioral guidance (tone, flow, conversational style) is expressed via Principles and 14 worked examples. Hard rules are reserved for factual constraints (price handling, eligibility, no minors) and safety boundaries (consent, prequalification gate). Examples demonstrate principles in action rather than prescribing behavior via rules. evidence (1)
|
| Check | Sec | Estado | Notas |
|---|---|---|---|
| 6a | 1.8 | PASS | NO_RESPONSE present (limits not checked, --no-db) |
| 6c | 3.x | PASS | 81 lines |
| 6d | 1.2 | PASS | 14 examples |
| 6e | 1.3 | N/A | no Principles section |
| 6g | PAT-006 | WARN | non-STEP-0 keyword handling |
| 6h | Anti-pat | PASS | no 'burbuja' |
| 6i | Anti-pat | PASS | no HTML comments |
| 6j | Anti-pat | FAIL | 48 em-dashes across 1 file(s); first: system.md:30 |
| 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 | PASS | URL forms consistent across files |
| 6q | PAT-001 | PASS | no instruction duplication into examples.md, keywords.json, program.json, resources.json |
| 6r | Keyword-drift | FAIL | 1 in keywords.json + recent digest, never named in system.md: ['action']; (3 stale drift entr(ies) ignored — no recent digest mention) |
| 6s | Hygiene | PASS | no expired dates in live SDK |
| 6t | Coverage | N/A | no Principles or Flujo section found in system.md |
| Fecha | Tema | Tags |
|---|---|---|
| 2026-07-13 | ai chatbot paused manychat switch | chatbot-status, tool-switch, cost-optimization |
| 2026-06-22 | minimal weekly activity content strategy | reporting, content-strategy, agent-activation |
| 2026-06-22 | minimal activity week content strategy alignment | campaign-planning, content-strategy, traffic-activation |
| 2026-06-22 | minimal weekly activity content planning | weekly-report, activity-monitoring, content-strategy |
| 2026-06-17 | subscription pause billing concern | billing, subscription, account-management |
| 2026-06-17 | subscription status and billing pause | billing, subscription, account-management |
| 2026-06-16 | billing suspension request inactive account | billing, account-management, service-suspension |
| 2026-06-15 | zero conversations second week cta missing | traffic, cta, content, funnel, critical |
| 2026-06-15 | zero traffic two weeks content cta gap | traffic, content, cta, funnel, critical |
| 2026-06-15 | zero conversations two weeks funnel stall | traffic, funnel, content-cta, critical |
| 2026-05-18 | weekly funnel report zero activity may 11 17 | automated-report, funnel-metrics, alert |
| 2026-05-18 | weekly report may 11 17 zero activity | reporting, funnel-metrics, alert |
| 2026-05-18 | critical zero activity week may 11 17 | funnel-metrics, alert, zero-activity |
| 2026-05-11 | weekly funnel report may 04 10 zero activity | reporting, funnel-metrics, alert |
| 2026-05-04 | low activity reactivation sprint proposal | activity, engagement, content-funnel, proposal |
| 2026-04-29 | instagram reconnection required agent setup | agent-setup, instagram, account-access, action-required |
| 2026-04-27 | weekly performance report apr20 26 | performance, metrics, bookings, funnel |
| 2026-04-27 | weekly survey request ninj checkin | survey, feedback, communication |
| 2026-04-22 | quarterly agent improvement call | agent-performance, optimization, feature-planning |
| 2026-04-22 | instagram reconnection request | account-management, platform-integration |
Sin pendientes en el checklist ✅
| Señal | Valor | Evidencia (cita del cliente) |
|---|---|---|
| Enojado con la plataforma | sí | “responses are shorter / So yes it is better” |
| Es queja | sí | — |
| Queja constructiva (accionable) | sí | — |
| Señal | Cita textual | Fecha |
|---|---|---|
| Queja recurrente #1: respuestas demasiado largas y robóticas (escalada crítica) | “None” | 2026-03-21 |
| Reconocimiento explícito de mejora tras v5 SDK — único feedback positivo directo del cliente | “responses are shorter / So yes it is better” | 2026-03-26 |
| Solicitud de kill-switch 'DEACTIVATE' — señal de falta de control sobre el agente | “None” | 2026-03-28 |
| Agente contactando amigos personales de Lloyd — pain point recurrente y sensible | “None” | 2026-03-28 |
| Pausa indefinida del chatbot AI — Lloyd vuelve a Manychat ($65/mes) por bajo volumen de conversaciones | “None” | 2026-07-13 |
| Links de Typeform muertos siendo enviados por el agente — Lloyd lo flaggeó; sin resolución en semanas | “None” | 2026-04-07 |
| Cero actividad durante semanas consecutivas (may 04–17, jun 08–22) sin diagnóstico resuelto | “None” | 2026-05-18 |
| Preocupación de billing: cliente pidió suspensión de facturación con cuenta inactiva | “None” | 2026-06-17 |
Sin cambios pendientes en el checklist ✅