AI use cases across every team

The problem

Every channel is flooded with generic AI slop — content that looks fine at a glance but has no real point of view, and no grounding in your product. No unique angles, nobody's checking search/AI visibility, and nobody's stress-testing copy against positioning. The result: lots of activity, very little compounding advantage.

Deliverables

Writers and marketers stop starting from a blank page or a generic prompt; they start from a shared product context and a distinct brand voice. Experiments ship much faster. Brand and content become visible to AI search engines.

Skills in this pack

• voice-extractor — Builds a reusable brand voice guide from samples so all AI output stays on-brand.
• positioning-basics — Locks down value proposition so every downstream channel has a single source of truth.
• icp-onboarding — Structured intake that defines who you're targeting, used by all outbound skills.
• angle-finder — Looks at a generic issue from many angles to find a truly unique positioning.

+18 more: landing-page-experiment, writing-thesis, ai-discoverability-audit, newsletter, ads & more

The problem

Reps spend a disproportionate share of their week on admin: writing follow-ups, updating the CRM, and rebuilding context before every call. Pipeline quality suffers because qualification is inconsistent and deal learnings never make it back into the process.

Deliverables

Reps reclaim meaningful selling time previously lost to admin. CRM data quality and pipeline accuracy improve because qualification is consistent and structured. Win/loss patterns surface and actually feed back into how the team sells.

Skills in this pack

• qualifying-leads — BANT/MEDDIC-style qualification so pipe reviews are grounded in reality, not gut feel.
• conversation-summarization — Extracts actions and risks from calls for follow-up and handoff.
• win-loss-reason-extraction — Analyses notes and emails to classify why deals are won or lost.
• pipeline-management — Defines stage hygiene and cadence so CRMs don't decay.

+10 more: objection-handling, discovery, closing, positive-reply-scoring & more

The problem

HR teams absorb an outsized share of repetitive operational load — onboarding checklists, policy questions, compliance monitoring — while key people leave for preventable reasons.

Deliverables

HR spends less time fielding repetitive questions and chasing checklists. Onboarding is consistent regardless of who runs it, compliance gaps surface before they escalate, and the team has the bandwidth to pay attention to all employees.

Skills in this pack

• hr-onboarding — Structures day-one and week-one experiences and checklists per role.
• employee-engagement — Designs engagement programs and interventions to maintain morale.
• hr-compliance — Keeps policies aligned with labour laws and flags gaps before they become incidents.
• hr-analytics — Builds people metrics and dashboards over existing HR data.

+5 more: conflict-resolution, compensation-benefits, job-description & more

The problem

Roadmaps are full, but conviction is thin. Features get prioritised because a loud stakeholder asked for them, specs are vague, and shipped work is hard to tie back to real user outcomes. The product org is busy, but not always learning.

Deliverables

The roadmap stops being a wish list and becomes a series of testable bets. PMs can say what a feature is supposed to change, how they'll know, and what they learned — and those learnings actually shape what comes next.

Skills in this pack

• problem-clarity — Checks if a proposed feature addresses a real user problem.
• outcome-definition — Reframes roadmaps around measurable outcomes vs output.
• assumption-mapper — Surfaces and ranks critical assumptions.
• experiment-design — Designs experiments with hypothesis, metrics, and decision rules.

+10 more: retention-drop-diagnoser, feature-adoption-analyzer, launch-readiness & more

The problem

Finance spends hours every month cleaning and reconciling numbers before anyone can even talk about what they mean. Different systems disagree, variances are massaged manually in spreadsheets, and by the time the story is clear, the window to act has passed.

Deliverables

Reporting cycles take less manual effort and produce more consistent output. Key metrics become shared knowledge. Cleaning becomes a defined, repeatable step instead of an endless, bespoke chore.

Skills in this pack

• finance-plugin-suite — Reconciliation, variance analysis, and reporting wired to your data sources.
• saas-metrics-coach — Guides operators through LTV, CAC, payback period, and unit economics.
• financial-analyst — Scaffolds DCF-style models and business analysis without a dedicated analyst.
• data-quality-audit — Checks data quality and anomalies before analysis.

+8 more: startup-analysis, saas-valuation-compression, finance-sentiment & more

The problem

Support quality varies by person and by shift. Agents spend the first stretch of every ticket rebuilding context. Frustrated users get inconsistent responses, escalations happen too late, and nobody has a clear view of what's actually breaking week over week.

Deliverables

Response quality stops depending on who is on shift. Escalations happen at the right moment rather than after a situation has already deteriorated. Recurring issues become visible patterns instead of invisible drains on team capacity.

Skills in this pack

• customer-support-plugin-suite — Triage, replies, escalations, and KB authoring across your support tooling.
• negative-sentiment-de-escalation — Structured response patterns for frustrated or angry users.
• handoff-detection — Identifies when an issue should escalate from AI to human agent.
• win-loss-reason-extraction — Extracts structured churn and cancellation reasons from conversations.

+3 more: conversation-summarization, claude-cs (custom support flows) & more

The problem

Institutional knowledge lives inside people's heads or in documents nobody updates. Every time someone leaves, a process disappears with them. Meeting decisions vanish into notes nobody reads, and new hires spend weeks asking questions that should have an answer somewhere.

Deliverables

Institutional knowledge survives team changes. New hires answer their own questions from day one instead of interrupting colleagues. Decision history exists and is searchable — not just in the memory of the person who was in the room.

Skills in this pack

• SOP-creator — Turns ad-hoc workflows into structured SOPs stored in your shared repo.
• meeting-transcript — Processes transcripts into decisions, notes, and follow-ups.
• enterprise-search — Lets Claude search and synthesise across your internal data sources.
• update-knowledge-base — Keeps docs synchronised as work progresses rather than drifting stale.

+9 more: braindump, weekly-checkin, team-brief, brand-onboarding & more

The problem

Data teams spend too much time cleaning messy data and reconciling dashboards that disagree. Every anomaly triggers a manual investigation that takes days. People field endless requests for one-off analysis instead of building the infrastructure that would make those requests unnecessary.

Deliverables

Cleaning becomes a repeatable step, not an endless manual scramble. Analysts spend less time reconciling numbers and more time explaining what changed and what to do about it. Anomalies surface faster, and investigations produce structured outputs instead of ad hoc Slack threads.

Skills in this pack

• programmatic-eda — Runs systematic exploratory data analysis for new datasets.
• metric-reconciliation — Forces KPI definitions to match across systems.
• data-quality-audit — Checks data quality and surfaces anomalies before they reach analysis.
• executive-summary-generator — Distils analysis output into stakeholder-ready summaries.

+16 more: funnel-analysis, cohort-analysis, ab-test-analysis, dbt skills, segmentation & more