Wesley Financial Group Marketing

Internal Tools

Tools for the WFG marketing team.

A small set of in-house apps that keep brand, content, and operations moving. More on the way.

Intel

Personas, channel performance, and funnel outcomes — who we reach, what converts, and where leads stall before they fund.

intel.wfg.marketing Open

Brand Guide

Logos, color palette, typography, voice, and visual standards: the single source of truth for anything WFG-branded.

brand.wfg.marketing Open

Reviews

Search, tag, and pull customer testimonials and review snippets for use across ads, landing pages, and social.

reviews.wfg.marketing Open

Testimonies

Browse and filter on-camera customer testimonial videos, with timecoded clips and ready-to-embed links for heroes, proof blocks, ads, and reels.

testimonies.wfg.marketing Open

Landing Page System

The shared system behind lp.wesleyfinancialgroup.com: how to take a page from idea to live, step by step, plus a look at the locked brand, tracking, and lead pipeline under the hood.

wfg.marketing/lp-system Open

MCP Server

Marketing Vault

The canonical source of truth for WFG brand, compliance, copy, audience, and product reference content. Connects to Claude as a read-only knowledge base; every WFG answer should reference it instead of guessing. Corrections route through GitHub Issues.

Tool Description
get_brand_guide The WFG brand and design system: color tokens, type scale, components, logo paths, voice rules, output-format notes, and build checklist. Auto-fires before any visual deliverable (HTML, PDF, slides, ads, social, email, infographic). Optional focus drills into colors, typography, buttons, tags-pills, cards, forms, shadows, layout, icons, logos, voice, stats, photography, output-formats, checklist, mcp, or locked-decisions.
list_brand_assets List the official WFG logo and lion-icon SVGs available through the MCP — name, filename, intended background, and use. Call this to pick the right variant (primary stacked vs. horizontal, light vs. reverse for dark backgrounds, one-color for single-ink), then pull the markup with get_brand_asset. Names mirror the CDN slugs in the brand guide (§10).
get_brand_asset Return an official WFG logo or lion-icon SVG as inline markup by friendly name (e.g. icon-primary for the gold lion, icon-reverse for dark backgrounds, logo-primary for the stacked lockup). The reliable way to get the real mark — it travels over the MCP, so it works even where the brand CDN is blocked. Never recreate, recolor, or redraw the lion; pull it here.
list_vault_files Index of all canonical Vault files with metadata (title, summary, tags, applies_to, status). Filter by category folder, by tags (topic keywords), or by applies_to (audience/use — note sales-internal marks closer/QS reference files not meant for marketing output). Start here when scoping any WFG task that touches brand, copy, compliance, personas, or stats.
read_vault_file Read a single Vault file by its Vault-relative path. Use after list_vault_files or search_vault has identified the right file. Optional section returns just one top-level (##) section instead of the whole file — pass a heading slug or text (e.g. "Red List"); on a miss it errors with the available section slugs.
search_vault Ripgrep across the Vault. Returns path, line number, snippet, and (for markdown hits) the enclosing section slug — pass that slug straight to read_vault_file's section param to read just the relevant section. Faster than listing-then-reading when you have a keyword, phrase, or stat to look up.
select_reviews Pull real, congruent WFG customer reviews (1,909 tagged) for any social-proof element: landing pages, ads, email, sales/video scripts, social. Filter by emotional_lever, funnel_stage, output_type, format_fit, audience_segment, objections_handled, and emotional_moments. Returns only wfg_positive reviews; resort names and legal/BBB/AARP references are flagged needs_redaction. Prefer this over search_vault for anything testimonial-related.
select_testimonials Select real WFG customer video testimonials (205 tagged) for on-camera proof: hero and proof videos, testimonial carousels, paid-ad clips, email embeds, and social reels. Filter by page_section_fit, emotional_lever, funnel_stage, format_fit, audience_segment, use_case, and more. Returns videos ranked by views with ready-made YouTube embed and deep-link URLs plus timecoded clips; resort/legal mentions flagged needs_redaction. Pair with select_reviews for written quotes on the same page.
get_canonical_stat Pull a canonical WFG metric (families helped, debt relief total, years in business, etc.) from the single source of truth. Always use this for numbers; never cite from prose or memory. Entries past the quarterly refresh window come back flagged stale: true with a warning — verify those with the user before publishing.
list_canonical_stats List all registered stat keys. Use when the user describes a metric without naming it, then pass the right key to get_canonical_stat.
get_data_interpretation_guide How to read WFG lead and revenue data. Fires before interpreting any marketing performance data — leads, lead sources, UTM attribution, revenue, CPL/CPA trends — whether pulled from Windsor.ai, Salesforce, GA4, or a spreadsheet. Raw numbers mislead without it: it carries the weekly lead goal (3,500) and alert thresholds, revenue tiers, UTM/attribution rules, per-source volume bands and quirks, landing-page naming conventions, fake-lead detection (~200–300/week are junk), holiday distortions, and which beliefs are verified vs. working hypotheses. Optional focus drills into scope, lead-goal, revenue, attribution, platforms, lead-sources, landing-pages, fake-leads, hypotheses, holidays, data-hygiene, or maintenance.
get_landing_page_system_guide How WFG landing pages get built and shipped. Fires before building, iterating, or reviewing any WFG landing page — and when anyone asks how to get a page live. lp.wesleyfinancialgroup.com is the recommended starting point for new pages (built-in attribution, GTM tracking, Chili Piper scheduling, compliance-gated review flow); the team makes the final platform call, and legacy platforms remain valid options — but pages built on the system must follow its locked patterns: lead capture only via the /info-kit/ link or <wfg-lead-form>, the no-staging PIN-gate review flow with approver-only go-live flips, tokens-only styling, and the Testing123 form-test convention. Optional focus drills into overview, lifecycle, building, lead-capture, automatic-layers, media, testing, where-things-live, or maintenance.
check_copy_compliance Pre-review compliance scan for finished WFG copy — run it when a draft is done and headed to review or publish, not while copy is still being drafted or explored. Checks against the canonical compliance guide: Red List terms, timeshare brand names, BBB/AARP references, definite inheritance phrasing, and numbers that look like WFG stats but don't match canonical values. Returns findings with severity (fix_before_publish / worth_a_look), the matched text and line, the guide rule, and the approved alternative, plus the pre-flight checklist. Optional channel (video | paid_ad | email | landing_page | social) adds channel-specific notes. Report-only: it never rewrites copy, and a clean scan is not an approval — human review is still final.
log_vault_update_request Open a GitHub Issue labeled "vault-update-request" to change the Vault — it's read-only through the MCP. Two request types: edit (default) for a fix that belongs in an existing file (exact current_text + suggested_text), or new_file to propose an entire new Vault file with a path and rationale; prefer an edit when in doubt. If an open request already targets the same file it returns possible_duplicate instead of filing again. The submitter is recorded automatically from the caller's @wesley.com login. Zach reviews.
list_vault_update_requests List the vault-update-request review queue: corrections flagged via log_vault_update_request that are pending triage. Filter by state (open by default) and optional severity. Returns compact rows; drill into one with get_vault_update_request.
get_vault_update_request Fetch one vault-update-request in full, with a true before/after diff read from the live file (whether it exists, how many times the current text matches, and whether approve will apply cleanly). For a new_file request, checks whether the proposed path is still available. Use to preview the exact change — or the proposed new file — before approving.
approve_vault_update_request Approve a request by opening a PR on a new branch — an edit applies current_text → suggested_text; a new_file request creates the proposed file (never overwriting an existing one). Never commits to main. A human reviews the diff and merges, which closes the Issue. Always preview with get_vault_update_request and confirm first.
deny_vault_update_request Deny a request: post the reason as a comment, label it wontfix, and close the Issue. A reason is required and is recorded so the submitter understands the decision.
vault_usage_report Vault usage analytics: which tools get called, which files get read, what people search for, and — highest-signal — searches that returned nothing (zero_hit_searches). Recurring zero-hit queries are questions the Vault can't answer yet: candidates for the knowledge-base-gaps file or a new-file request via log_vault_update_request. Optional days sets the trailing window (default 7); counters persist on the deployed server.
health Liveness check + build metadata (inventory present, stats count, build time).