Marchenko-sales

b2b-lead-contacts

Deep search for named decision-makers at a specific company: names and roles of the people who decide on buying your product, their LinkedIn and personal Telegram, the corporate email pattern and probable personal addresses with confidence levels. An expensive step — run ONLY when a generic channel (info@/partners@) is not enough or the user asks directly. Company context (legal entity, reputation, buying signals) comes from company-context — run it first. Skill body is in Russian. Triggers (EN): 'find the decision-maker', 'who is the founder', 'reach the CEO', 'find a personal email or Telegram'. Triggers (RU): «найди ЛПР», «кто основатель», «выйди на CEO», «найди личный email/телеграм».

cold-outreach

Cold outreach for B2B: personalized first messages — email, Telegram, LinkedIn (accepted connections) and generic inboxes (info@) — for every contact from the dossier, plus a 4-touch follow-up cadence. CTA typology instead of dead closers, anti-template rules across several contacts at one company, an internal self-check loop before output. Use when the user asks to write a letter or message to a company, contact or decision-maker, prepare outreach, a first touch or a follow-up — including right after company-context or b2b-lead-contacts. Skill body is in Russian. Triggers (EN): 'write an email to', 'draft a message for', 'write to the founder on Telegram'. Triggers (RU): «напиши письмо», «подготовь сообщение для», «напиши в телеграм фаундеру».

company-context

Pre-outreach company dossier for B2B sales — so the first message is specific instead of 'we'd like to offer cooperation'. Collects: legal entity and jurisdiction, reputation, public contact channels (info@, partners@, Telegram, forms), the current vendor in your category, and buying signals with a hook for your offer. Does NOT search for named decision-makers or personal emails — that is a separate, more expensive step. Skill body is in Russian. Triggers (EN): 'build a dossier on company X', 'is this our lead?', 'what do they use now?', 'what angle to approach with'. Triggers (RU): «собери контекст по компании», «что за компания», «наш ли это клиент», «с каким поводом заходить», «чем они пользуются сейчас».

qualification-objections

Lead qualification and objection handling for B2B sales, on calls and in writing. Sorts a lead into three buckets (ICP / referral to a partner / decline), gives discovery questions for the first contact, and principles for answering objections about price, timing, reliability, 'we already have a vendor' or requests for things you don't do. Skill body is in Russian. Triggers (EN): 'is this our lead?', 'qualify this client', 'what to ask the lead', 'how to answer this objection', 'the client says it is too expensive'. Triggers (RU): «наш ли это лид», «квалифицируй клиента», «что спросить у лида», «как ответить на возражение», «клиент говорит дорого».

segment-company-search

Segment prospecting for B2B sales: finds candidate companies in a given niche or hypothesis (e.g. 'medical tourism', 'online English schools', 'car rental services'), verifies each against two strict criteria (segment fit + your ICP signal) and saves a prioritized xlsx (priority signal × company scale). Use when the input is a segment description, NOT a specific company. Output feeds company-context. Skill body is in Russian. Triggers (EN): 'find 30 companies in niche X', 'list candidates in segment Y'. Triggers (RU): «накидай компаний в сфере X», «найди 30 компаний ниши Y», «список площадок в сегменте Z».

stop-slop-ru

Removes recognizable AI-writing patterns from Russian-language text: throat-clearing openers, false contrasts, bureaucratese, dead cold-email closers. Use when writing, editing or final-checking any Russian text — emails, messages, posts — before sending. Hard mode for cold outreach. Russian adaptation of the MIT-licensed stop-slop by Hardik Pandya. Skill body is in Russian. Triggers (EN): 'remove AI clichés', 'clean up this Russian text', 'does this sound like a bot?'. Triggers (RU): «убери ИИ-штампы», «почисти текст», «звучит как нейросеть», «проверь перед отправкой».