HBICHealth & Beauty Institutional Chaos · Mission Control
Today Pulse AI Maturity Skills Setup
2026-10-01 · refreshed 4:47 PM CT

The short version

Your assistant has three layers stacked on top of each other: rules (the rulebook — what to do, what not to do, how to behave), skills (specific tasks it knows how to perform — write a 5/15, evaluate a deal, run a BigQuery query), and memory (everything it's learned about you and your work over time).

On top of that, 36 jobs run automatically in the background — refreshing this dashboard at 5pm Central, building the morning briefing, pulling refund data Mondays at 10am, and so on. You don't trigger these; they just run.

8 hooks watch every message you send — they catch corrections you make (so future Claude learns from them), spot when a session is ending (so work gets logged), and remind Claude to invoke the right skill before answering certain prompts.

The point of this page: a single place to see what's actually wired in your setup — so when something feels off or a routine seems broken, you can spot it instead of guessing.

12Rules
91Skills
162Memory files
36Scheduled jobs
8Hooks

Rulebook

how Claude is told to behave — loaded into every conversation
4
Global rules (apply everywhere)
Personal behavior rules loaded in every session, regardless of project.
10x-dashboard-tracking.md10x-value-impact.mdhonor-batch-authorization.mdskip-deprecated-and-redundant-skills.md
8
CoS workspace rules
How to work inside the CoS workspace — skill enforcement, scope discipline, verification gates, snapshot reuse.
asana-api-over-mcp.mdcanonical-skill-over-precedent.mdceo-filter.mdcheck-snapshot-before-pull.mdscope-diff-pregate.mdskill-enforcement.mdteam.mdverify-rendered-output.md

Skills

specific things Claude can do — invoked by name or matching a trigger phrase
62
Global skills (~/.claude/skills/)
Available in every project — domain skills, file handling, dashboards, voice.
weekly-515Generate Katelyn's weekly 5/15 progress summary, pull canonical numbers from her own dashboards, draft in her voice, and publish to GIQ for her audit.
virtual-dusanVirtual Dusan is a CEO persona skill that channels a bootstrapped, AI-first, value-creation-obsessed entrepreneur and operator.
describe-my-setupGenerate a one-page HTML overview of this Claude setup — the rulebook (CLAUDE.md + rules), skills, scheduled routines, memory, and any automations — with a plain-English summary at the top.
system-auditMonthly check-up of this Claude setup, in two passes — behavioral drift (where the user keeps correcting Claude, turned into proposed rules) and setup hygiene (are rules/skills at the right level, overlapping, or stale; does the config still match reality).
groupon-iq-reportsUpload, manage, and share HTML reports via Groupon IQ REST API.
ai-showcase-digestWeekly digest of Groupon's AI Showcase Google Chat space, filtered and summarized for Katelyn Condon (HBW Category Management).
10x-coachCoaching agent that tracks work patterns, measures against five CEO-defined metrics, identifies gaps, and provides actionable coaching.
account-health-reportProduce a rep-facing Account Health Review for a Groupon merchant account that has multiple active deals — financial trend, per-deal breakdown, refund root-cause from SaltedCx Customer Logs, deal expansion opportunities, and action recommendations for the Account Manager and their manager.
agnes-data-queryingUse when querying any data in Agnes — discovery first, estimate before fetch, materialize scoped subsets locally
analytical-context-binderBinds the right analytical context to the user's question by fetching relevant data catalog context (data products, metrics, glossary dimensions) and identifying which skills from Git to use.
asana-515-checkCheck Asana 5/15 weekly submissions for a team.
asana-hygiene-checkAudit Asana subtask hygiene for a list of parent tasks.
ask-data-bigqueryExecute BigQuery SQL queries for data analysis, exploration, and schema inspection via Ask Data MCP.
ask-data-openmetadataDiscover and explore Groupon data assets using the Ask Data MCP OpenMetadata tools (search_metadata, get_metadata_entity).
ask-data-tableauQuery and analyze Groupon data through Tableau datasources via Ask Data MCP.
automation-finderAnalyzes Groupon processes for automation potential.
brainstormingYou MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior.
category-driven-salesBuild a complete category sales drive — interactive HTML seller guide (unit economics, deal calculator, CAC comparison, targeted motion emails + call, objection handling, competitive shark list) plus a 4-week Asana sprint plan.
category-pulseWeekly market-level performance dashboard for LHR (Laser Hair Removal) in a specific Groupon DMA/division.
consolidate-memoryReflective pass over your memory files — merge duplicates, fix stale facts, prune the index.
cowork-onboardingInteractive onboarding assistant for new Groupon Finance Cowork users.
daily-morning-briefingDaily morning briefing for Katelyn Condon — calendar + categorized inbox (with mandatory CEO treatment) + Asana — posted to Google Chat via webhook.
daily-morning-briefing-v2Katelyn's opinionated morning operating brief.
deal-analystGroupon deal data domain skill.
deal-competitive-benchINTERNAL PIPELINE HELPER — called by `deal-health-report` with `--with-competitors`.
deal-health-reportOrchestrates a full deal health evaluation by dispatching parallel subagents across content analysis, performance metrics, and (optionally) competitive benchmarking.
deal-mriGenerate a full "Deal MRI" diagnostic HTML report for any Groupon deal.
deal-page-evaluatorINTERNAL PIPELINE HELPER — called by `deal-health-report`.
deal-page-extractorINTERNAL PIPELINE HELPER — called by `deal-health-report` and `deal-page-evaluator`.
deal-performanceINTERNAL PIPELINE HELPER — called by `deal-health-report`.
deal-resolverINTERNAL PIPELINE HELPER — called by `deal-health-report`.
devil-advocateStructured devil's advocate analysis — challenges any idea, decision, strategy, or proposal by steelmanning it first, then stress-testing it from multiple angles.
fact-checkerSystematic fact verification and misinformation identification using evidence-based analysis.
fai-agnes-migrationUse when migrating a scheduled job or dashboard off the dead direct-BigQuery path onto Agnes (the FoundryAI CLI), or when an existing Agnes query is failing with a "Table X is not in your stack" / "Syntax error at CASE" / "Only single SELECT queries are allowed" / "HTTP 429 concurrent_slot_exceeded" / column-shifted CSV output error.
file-organizerIntelligently organizes your files and folders across your computer by understanding context, finding duplicates, suggesting better structures, and automating cleanup tasks.
find-skillsHelps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities.
freshness-sentinelDaily M-F 09:45 watchdog that checks every scheduled job on this Mac and posts a single gchat digest listing anything stale or errored.
gchat-notifyPost a message to a Google Chat space via webhook (curl).
groupon-category-playbooksGroupon business domain knowledge - deal structures, category rules, take rates, merchant lifecycle, compliance.
groupon-clean-themeMinimalist Groupon-branded theme for all output types (HTML, PPTX, DOCX, XLSX, PDF).
hbw-mbr-reviewerReview and validate the HBW (Health, Beauty & Wellness) Monthly Business Review before it reaches Dusan or the CEO.
incident-quality-analyzerThoroughly analyze incident response quality from war room transcripts, meeting notes, or incident records.
jiri-scorecardScore a 5/15 draft (or any weekly status update) against the Jiri-Ready rubric and produce a "🔍 WEAK SPOTS TO TIGHTEN BEFORE SENDING" punch list.
katelyn-voiceWrite in Katelyn Condon's authentic voice and pass every draft through the 4-lens audit — fabrication check, AI-slop scrub, authorship discipline, voice/register match.
lead-generationBuild a Sales-Ops-ready lead list (CSV) for Groupon HBW (Health/Beauty/ Wellness) merchant prospects.
legitscript-badge-detectorDetect whether a LegitScript certification badge is visible on a merchant's website.
market-manager-skillGroupon Market Manager (MM) operations and analytics skill.
meeting-notes-to-asanaAutomated meeting notes from Google Calendar events with Gemini transcripts into Asana tasks.
merchant-analystGroupon Supply Analyst skill for the merchant data domain.
option-title-normalizerINTERNAL PIPELINE HELPER — called by `deal-health-report` as Step 2, immediately after deal-page-extractor output is available.
pii-access-control-testRun a read-only data access control test that checks whether this user's Google/BigQuery credentials can reach restricted customer data tables via Claude's BigQuery connector, and record a PASS/FAIL result against the runner's work email (the test content itself is anonymized - no customer data is ever read or logged).
refund-contact-insightsOne-stop Refunds & Contacts analyzer for Groupon CommOps and Supply / Merchant Ops.
refund-review-scaffoldOrchestrate a full L2/L3/L4/PDS refund review for Katelyn's HBW categories — runs the refund-contact-insights analysis with majority-of-scope filter, publishes an HTML report to GIQ with per-deal row anchors, creates a parent Asana task in HBW Control Center (or attaches to an existing one), and bulk-creates top-N per-deal subtasks pre-populated with the locked template (KPIs, refund reasons, verbatims, SF Opportunity + SF Account links, GIQ deep-link, live deal hyperlink, owner + manager).
report-templateGenerate self-contained HTML reports and articles with consistent branding, data-driven components, and mobile responsiveness.
rx-merchant-prescreenPre-screen a prospective Rx / telehealth / medical MERCHANT for the Groupon prescription program — a risk verdict on a business before Groupon advertises it.
salesforce-genericConnect to Salesforce and run SOQL queries or pull report data by ID.
setup-coworkGuided Cowork setup — install a matching plugin, try a skill, connect tools.
shopping-list-analystGroupon Shopping List (SL) performance analysis skill.
skills-mapGenerate a self-contained interactive HTML map of every skill available in this Claude setup, clustered by intent (intent hubs -> sub-clusters -> skills, with cross-cut edges).
ue-metrics-comparatorExplain how a metric is constructed in Unit Economics (UE Keboola pipeline) and optionally compare it to the Near-Real-Time (RT) financial metrics pipeline.
unit-economics-guide-generatorGenerate interactive Unit Economics & Seller Guide HTML pages from HBW Category Playbooks.
wbr-hbw-refreshRun the weekly HBW L2 WBR dashboard refresh pipeline — pulls fresh BQ data, builds the main Global dashboard AND the NA Deep-Dive Companion, publishes BOTH to GIQ, pings Katelyn on GChat.
+ 56 more · click to expand
29
CoS workspace skills
Project-tuned skills layered on top of the global set — specialized for your workflows.
cos-welcomeSession startup orientation — interactive project picker and workspace navigation.
new-projectCreate a new project in the CoS workspace.
asanaUse when bulk-downloading Asana tasks, syncing Asana data to local DB, or when MCP would require too many API calls that bloat context window.
bigqueryExecute BigQuery SQL queries for data analysis, exploration, and schema inspection.
bigquery-analystWrite and run BigQuery SQL queries for data analysis.
ceo-voiceWrite in Dusan's authentic CEO voice — direct, intellectually passionate, no corporate speak
compress-outputForces agent to compress its final output before returning to Leader.
compress-reviewReviewer-specific output compression — preserves findings detail while staying concise.
data-catalog-assistantAnswer user queries about data assets using OpenMetadata Data Catalog via MCP Server.
decision-docCreate structured decision documents with options analysis, trade-offs, and revisit triggers.
deep-researchExecute autonomous multi-step research using Google Gemini Deep Research Agent.
docxUse this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files).
frameworksLoad strategic thinking frameworks during brainstorming, planning, or strategy work.
groupon-enterprise-architectUse when designing, reviewing, or deciding on architecture for Groupon systems.
local-mlRun local ML models for embedding, evaluation, and clustering.
migrate-projectMigrate an existing project to the enhanced folder structure and CLAUDE.md template.
pdfUse this skill whenever the user wants to do anything with PDF files.
pptxUse this skill any time a .pptx file is involved in any way — as input, output, or both.
publish—
refresh-registryScan projects/ filesystem and reconcile docs/projects.md with actual project folders.
review-panelRun internal multi-perspective review using 4 persona subagents (engineer, executive, skeptic, user-voice).
sales-enablementWhen the user wants to create sales collateral, pitch decks, one-pagers, objection handling docs, or demo scripts.
salesforceConnect to Salesforce and run SOQL queries or pull report data by ID.
saltedcx-queryingQuery and analyze SaltedCX B2B call transcript data from BigQuery.
second-opinionGet an external AI review of plans, analysis, decisions, code, or any significant output using available CLI tools (claude, codex, gemini)
skill-creatorCreate new skills, modify and improve existing skills, and measure skill performance.
swarmPhased workflow orchestration.
watchlist-managerManage Katelyn's lightweight deal watchlist — add a deal to watch, mark one resolved, reactivate, list what's being watched, or run the deep-check that pulls fresh metrics.
xlsxUse this skill any time a spreadsheet file is the primary input or output.
+ 27 more · click to expand

Memory

what Claude has learned about you and your patterns — auto-loaded into every CoS session
162
Memory files
Indexed in MEMORY.md. Feedback rules, project canon, references, user profile. MEMORY.md index is 28.4KB — at the edge of the 24.4KB loading cap; trim recommended.
feedback_*.md (137) project_*.md (5) reference_*.md (15) user_*.md (1) grouped by type, indexed in MEMORY.md

Scheduled routines

jobs that run automatically — no one needs to trigger them
7
HBIC dashboard jobs (com.hbic.*)
Keep this dashboard fresh — light refreshes through the day, calendar/Gmail sync, refund briefing, pulse data.
email-calendar-refreshgrowth-pool-refreshheavy-refreshlight-refreshnorthstar-reminderpulse-refreshrefund-briefing
23
Personal & category jobs (com.kcondon.*)
Friday 5/15, daily morning briefing, MBR monthly build, AI Showcase digest, LHR pending reneg, Botox floor sweep, enterprise tracker, memory loop.
ai-showcase-digestbotox-floor-friday-refreshbotox-floor-sweepcloudsdk-python-setenvdaily-briefing-feedback-scannerdaily-morning-briefingdaily-morning-briefing-v2franchise-tracker-monthly-refreshfreshness-sentinelfriday-515hbw-enterprise-tracker-refreshhbw-industry-intel-weeklyhbw-mbr-monthly-buildhbw-search-trends-monthlyinjectables-market-weeklymeeting-vault-ingestmemory-loopmonthly-auditmonthly-rep-productivityq4-hbw-october-checkwatchlist-dailywatchlist-weeklywbr-hbw-refresh+ 16 more · click to expand
6
Groupon work jobs (com.groupon.*)
Weekly refreshes for Groupon-wide trackers — CM impact, HBW portfolio adoption, Rx weekly performance.
cm-impact-refreshglp1-weekly-sweephbw-portfolio-adoption-refreshrx-sf-reconciliationrx-weekly-refreshweekly-515-sweep

Hooks & connectors

the wiring between Claude and the outside world — fires on every message
8
Hooks (watch every message)
Catch corrections so you don't have to repeat them, log work at session end, remind Claude to invoke the right skill before answering.
authorization-detectorcorrection-detectorcorrection-loggerkatelyn-voice-enforcekatelyn-voice-mark-loadedkatelyn-voice-trigger-detectorsession-end-detectorskill-check-reminder
4
MCP connectors (live data)
Connect Claude to live external systems — Tableau, BigQuery, OpenMetadata, library docs, browser automation.
DataCatalog (workspace) ask-data (Tableau + BQ + OpenMetadata) context7 (library docs) playwright (browser)