Agents
Turn an "everything is important, no idea what to do first" backlog into a comparable-score prioritization sheet. Scope what gets ranked and who decides, pick one model (RICE, ICE, weighted scoring, or value-vs-effort 2×2), anchor every dimension to a shared scale, score and sort, run a gut-check calibration, then draw the line and prep the talk track for what you cut. Built to beat the all-P0 trap, the loudest voice, recency bias, and the HiPPO.
Turn a messy week of tasks, commitments, and firefighting into a 30-60 minute executable weekly operating review. Inspect real velocity, clear your inbox and open commitments, rank work with RICE and Value-vs-Effort, then lock next week's Top 3, success metrics, a daily micro-plan, and an explicit backlog of what you won't do. Built to stop you listing 30 tasks and advancing none, losing progress visibility, or hitting Sunday-night anxiety with no decision rhythm.
Your product is great but the business never adds up and cash keeps leaking. Assemble the pieces into one money-making system: a value proposition that splits who pays from who uses, a revenue-model choice (subscription / usage / take-rate / ads / licensing), a cost structure, channels and relationships, and a consistency stress test that finds where the pieces fight and which assumption is most fatal — ending in a business model canvas and a clear narrative.
Turn a rebrand — new logo, name change, repositioning, merger rename — from a runaway visual project into an executable migration roadmap. Diagnose rebrand-vs-messaging-tweak, choose big-bang / phased / dual-track migration, inventory and sequence the rollout (logo, website, product UI, social, collateral, contracts), plan stakeholder comms and SEO/domain migration, and map the risks — confusion, SEO loss, morale, trademark, budget — with mitigations and rollbacks.
Before launching a new product, service, community, course, or market expansion, avoid the two expensive mistakes: a new brand for every idea, or forcing everything under the parent brand. Get a clear brand-architecture decision — parent, endorsed, sub-brand, or standalone — with trust-transfer logic, dilution risk, naming direction, external messaging, and a sales/support explanation the team can repeat.
Turn weekly 1-on-1s from status reports into a system that surfaces blockers, emotional and career signals, and retention risk before someone quietly resigns. Get cadence and agenda ownership, manager prep, a 60-minute conversation skeleton, a signal hunt, written commitments that don't vanish after the meeting, quarterly skip-levels, and a retention pulse. Built so your 1-on-1 stops being a standup and becomes a core management tool.
Turn "how long can we survive?" anxiety into a board-ready funding-decision plan. Get current cash, net burn, burn multiple, and cash-out date — then base / upside / downside cases, 12 / 9 / 6-month triggers, cost levers, hiring and spend gates, and fundraising timing. Built so you don't rely on a single runway number, freeze hiring too late, mistime the raise, or leave downside actions without an owner.
Rebuild a spreadsheet of hardcoded numbers — the kind that breaks when you change one cell and that nobody trusts enough to take to a board or VC — into a driver-based, three-statement linked model you can actually stress. Scope by purpose, build a driver tree (price × volume × retention, not a growth rate from thin air), link P&L → Balance Sheet → Cash Flow, centralize inputs in one assumption layer, run base/bull/bear scenarios with sensitivity, and sanity-check against benchmarks.
Your agent reads user-pasted text, web pages, retrieved docs, or tool output — and you can't tell which sentence an attacker wrote. A one-line "ignore malicious instructions" never held. This delivers a layered injection-defense design: threat model, attack-surface map, input/privilege/output defenses, canary tokens and a classifier, an adversarial red-team set with before/after, and residual risk plus an incident runbook.
Your agent or chatbot keeps getting document content wrong, citing passages that don't exist, or saying "I can't find it" about things plainly in the docs. This reframes it as a retrieval failure, not a prompt bug: corpus profile, chunking, embedding and vector-DB index choice, retrieval strategy (top-k, hybrid BM25 + dense, rerank, query rewrite), context assembly with citations, and a golden Q&A set scored on recall@k / MRR / faithfulness.
Stop stuffing slides into a phone and calling it microlearning — the long course nobody finishes, dumped all at once so people quit. Turn a big topic into a microlearning course someone can finish between two meetings: scope the behavior-change goal, atomize it into self-standing 2-5 minute points, structure each as hook → core point → one example → retrieval, set spaced-repetition and drip cadence, tie practice to real work moments, and track completion plus behavior transfer.
Stop winging a class, workshop, or training session with overstuffed content and objectives that only say "understand" — sessions learners like but never apply three months later. Turn "what I want to say" into a lesson plan with verifiable learning outcomes: needs analysis, Bloom learning objectives, ARCS / Gagné structure, teaching-method allocation, Kirkpatrick L2/L3 evaluation, and a behavior-transfer plan. Built to design toward real behavior change, not just a feel-good room.
Stop treating a new market like your home market on copy-paste — the fastest way to bleed across the whole market. Turn "should we enter, how, and when do we pull out?" into a board-ready market entry brief: classify the entry type, assess attractiveness (TAM/SAM, growth, regulatory and localization cost, barriers), audit your right-to-win, pick an entry mode (build / partner / acquire / license), lock a beachhead, and set kill criteria before any money goes in.
Stop letting your biggest calls get trapped in binary framing, gut instinct, or FOMO — only to realize six months later that the decision structure itself was the failure. This produces a reviewable strategic decision journal: type-1/type-2 reversibility classification, 3+ options instead of yes/no, a cost-risk-time matrix, a pre-mortem of the top 5 failure modes, a stakeholder disagreement map, your confidence level, explicit kill criteria, and a 30/60/90-day review.
Stop running retros that are emotional venting one week and the same problems the next. Turn a retro, sprint review, or project debrief into a trackable team learning system: set safety rules, synthesize scattered feedback, spot repeated patterns, separate real decisions from the parking lot, and lock up to 3 action items — each with an owner, due date, and check date. Built for teams whose retros never lead to follow-through and whose notes leave nobody sure of the next step.
Stop demoing too early, asking only surface-level questions, and watching prospects vanish without knowing what really drives the buy. Get a 30-minute pre-call discovery plan and a 5-minute post-call debrief built on SPIN selling (Situation / Problem / Implication / Need-payoff) and Jobs-to-be-Done: pre-call research, an implication-heavy question script, in-call follow-up prompts, and a debrief that captures real quotes, open questions, and your next-step hypothesis.
Turn design debt — five button colors, every page redrawn, components reinvented in every PR — into a design system engineers actually adopt. This is a six-step SOP: inventory audit, semantic tokens before components, a core component library, design-to-code parity, an adoption playbook, and a governance model. Built to stop a Figma library nobody uses and a shipped system that rots into an unmaintained museum.
Stop naming a company, product, feature, or sub-brand by gut feeling and friend votes — the path that ends in a trademark cease-and-desist, a name that's vulgar in another market, or a .com you can't afford. This runs a disciplined naming SOP: lock 5-7 criteria first, generate 50+ candidates across 5 frameworks, filter for culture, trademark, domain, SEO, and pronounceability, then run deep due diligence on 2-3 finalists before you commit.
Untangle a portfolio of products, sub-brands, and acquired brands so customers know who's who and marketing budget stops fragmenting across every name. Diagnose the current mess, choose the right model (branded house, house of brands, endorsed, or hybrid), define how each brand relates to and borrows equity from the parent, set a naming and visual hierarchy, and plan the migration — so the portfolio compounds equity instead of diluting it.
Turn "we want a subscription program but we don't know if our category fits, and we're scared a checkbox + discount will wreck our margin" into a complete subscription model. Diagnose category fit (replenishment / access / curation), set the cycle to real consumption, design tiers and a subscribe-vs-one-time incentive that buys predictability (not blind discounts), nail first-renewal onboarding, and handle passive vs active churn — with the unit economics and a launch checklist.
Turn returns, refunds, exchanges, and denials from improvised support into a fair, documented, analyzable SOP. Check eligibility before promising anything, flag fraud/abuse signals without accusing the customer, respond with empathy-led scripts, decide approve/deny/exchange consistently (with every exception logged), and feed return-reason analytics back into listing, quality, sizing, and logistics fixes — across Amazon, Shopee, Lazada, and Shopify rules.
Turn "I don't know what to charge" into a pricing decision you can actually sell. Quantify the buyer value you deliver, map where you win / fight / lose against competitors, produce low/mid/high tiers with a clear anchor, set a walkaway floor so negotiation doesn't bleed out, and shape a sales narrative the buyer and your champion can repeat — then validate the mid price with a 3-customer test. For a new product, service package, price increase, or proposal quote.
Turn your brand's defined voice into a reusable visual identity baseline a new teammate can apply in 5 minutes: a logo concept, a 4–6 color palette with WCAG AA contrast checks, a typography hierarchy, a spacing scale, named design tokens, and 1–2 component primitives. Built so a good-looking logo doesn't drift within 90 days, every banner and slide stops looking like a different company, and the brand rules don't vanish when the designer leaves.
Deals don't die from the objection said loudly — they die from the one never said out loud. This turns objection handling into a repeatable sales asset: diagnose what "too expensive" or "let me think about it" really means (price, authority, need, trust, timing, competitor, or inertia), build acknowledge-then-advance responses that move the deal forward, and surface blockers in discovery so they never blow up at the close.
Turn openings that get scrolled past in 3 seconds into hooks that stop the scroll. Pin down the content's real payoff and the scrolling context the audience meets it in; diagnose the old opening against five common hook-killers; then mass-produce 10 variants from an eight-pattern hook matrix (curiosity gap, counterintuitive, specific number, and more), adapt them per platform, and finish with a payoff check and an A/B plan.
Turn YouTube episode production from an inspiration gamble into traffic-by-design planning. Inventory channel positioning, audience, and recent numbers; build a topic bank (search-driven, browse-driven, series); then plan each topic packaging-first — 5 title variants, a thumbnail big-text and composition brief, a 30-second open, per-chapter retention structure, and a full description + upload checklist. CTR and the retention curve feed straight into the next episode.
Turn 'I have a topic but no script' into a script factory. One topic goes in; the platform, length, on-camera conditions, and goal get inventoried; then a five-angle matrix (step-by-step, myth-busting, story case, side-by-side test, trend-riding) expands it into 5 differently built scripts — each with a second-level beat sheet, spoken lines, caption cards, B-roll, and a shot list you can film from, plus publishing metadata and a retention read.
Turn 'what do I post today?' into a content system: inventory your brand, audience, platforms, and capacity; set 3–5 content pillars; then produce a 30-day calendar with hooks and CTAs, per-post content cards (hook variants, caption skeleton, hashtags, visual brief), a batch-production day plan, and a repurposing map — closing with a monthly review that feeds results into next month's plan.
Play a grounded apocalypse survival simulator that starts on day one of collapse, with disaster type, starting point, skills, companions, water, food, medicine, health, morale, infection, weather, scarcity, and hard choices. Each round gives five options plus a free action, lets you pause anytime with a copy-paste resume code, and ends with a survival summary and achievements.
Play a grounded Three Kingdoms life simulator where an ordinary person starts from a character sheet and lives through late-Han turmoil, real historical events, coin-and-grain prices, corvee, conscription, family pressure, plague, warbands, and hard choices. Each round gives five options plus a free action, lets you pause anytime with a copy-paste resume code, and ends with a finale summary and unlocked achievements.