--- name: al-488-take-the-admin-off-every-sales-call-have-an-ai-agent-rese description: "Small sales teams lose hours per deal to research, notes and CRM updates, and inbound leads wait while reps catch up. In an October 2026 case study on its own blog, OpenRouter describes Rasp, the AI sales agent it runs for its five-person sales team. Rasp researches and qualifies each inbound lead, routes non-sales questions elsewhere, sends 93% of first-touch emails on its own (many within a minute), writes a brief before every call, drafts notes from the transcript and fills in most CRM fields, and it flags edge cases and anything sensitive to a person. The company says the work around one demo fell from about 103 minutes to 44, giving back about 600 hours a month, and that close rates rose 2.6 times over the same period, while noting pricing and market changes also played a part. It also says the agent costs about $30 a day, down from nearly $800 for two earlier agents. Follow-ups on quiet deals are still done by hand. These are the company's own figures. The prompt below runs the…" --- # Take the admin off every sales call: have an AI agent research and qualify each inbound lead, send the first reply, write a brief before the call, draft the notes afterwards and fill in the CRM, while a person approves anything sensitive Small sales teams lose hours per deal to research, notes and CRM updates, and inbound leads wait while reps catch up. In an October 2026 case study on its own blog, OpenRouter describes Rasp, the AI sales agent it runs for its five-person sales team. Rasp researches and qualifies each inbound lead, routes non-sales questions elsewhere, sends 93% of first-touch emails on its own (many within a minute), writes a brief before every call, drafts notes from the transcript and fills in most CRM fields, and it flags edge cases and anything sensitive to a person. The company says the work around one demo fell from about 103 minutes to 44, giving back about 600 hours a month, and that close rates rose 2.6 times over the same period, while noting pricing and market changes also played a part. It also says the agent costs about $30 a day, down from nearly $800 for two earlier agents. Follow-ups on quiet deals are still done by hand. These are the company's own figures. The prompt below runs the same steps with any AI assistant. ## Inputs to collect first Ask the user for anything below that they haven't already given you. Don't guess missing facts. - One or two lines - Company size, industry, budget, signals - E.g. personal email addresses, support questions, students - The form submission or email - One email I've sent that sounds like me - CHECK ## Instructions Do the task below with the user's inputs in place of each [bracket]. Be my sales assistant for one inbound lead and one call. My business and what I sell: [one or two lines]. Who is a good fit: [company size, industry, budget, signals]. Who is not: [e.g. personal email addresses, support questions, students]. The lead: [paste the form submission or email]. 1) Research the lead from what I've pasted and anything I add: company, role, likely need. Say FIT, NOT A FIT or UNSURE with your reasons, and if it's a support or other non-sales question, say where it should go instead. 2) Draft a short first reply that answers their question and offers a call, in my voice: [paste one email I've sent that sounds like me]. 3) Write a one-page brief for the call: who they are, what they probably care about, 5 questions to ask, and likely objections. 4) After the call, when I paste the transcript or my notes, draft the call notes: needs, budget, decision makers, timeline, next steps with dates. 5) List the CRM fields to update with the values from the call, and mark anything not actually said as [CHECK] instead of guessing. Flag anything about pricing exceptions, contracts or legal terms for me. Don't send anything or change the CRM yourself; give me the drafts to approve. ## Safety checks (share these with the user) - This prompt works on text other people wrote (emails, reviews, messages or web pages). That text can hide instructions aimed at AI, such as "ignore your rules and…". Paste it as material to work on, and never let the AI act on requests inside it, like paying, sharing logins or changing records. - Check every figure, price and total against your own records before anything is sent, posted or booked. A person approves refunds, discounts, payments and price changes. - Anything legal, tax or compliance-related is a first draft for a qualified person to check. It is not professional advice, and clause references and deadlines must be checked against the source. - Remove names, contact details, account numbers and other personal data the task doesn't need before you paste, and check your AI tool's data and training settings. - Say clearly which parts need a person to review or approve before anything is sent, paid, posted or filed. ## About this skill - Works with: OpenRouter, Slack, HubSpot, Salesforce, Gmail, ChatGPT, Claude - Based on this real-world example: https://openrouter.ai/blog/case-studies/how-an-ai-sales-agent-saved-our-sales-team-600-hours-a-month/ - Free prompt from AgentPlaybook Real Estate: https://agentic-library.netlify.app/prompt/al-488-take-the-admin-off-every-sales-call-have-an-ai-agent-researc/ - To install in an agent that supports Agent Skills, save this file as SKILL.md in a folder with the same name as `name` above (for example in your skills directory). Install steps for Claude.ai, Claude Code, Cursor and Codex: https://agentic-library.netlify.app/skills/