zRev AI

Returns a plain-text profile of zRev AI: what the firm does, the companies it is the best fit for, who it is not for, its six service lines with a link to each, and contact details. Use this first when a user asks who zRev is, what it offers, or whether it suits their company. Read-only, static content, no arguments, no authentication, no rate limit.

Takes no parameters.

Returns zRev's typical engagement results as plain text: 2x pipeline velocity, 20% lower customer acquisition cost, 10 hours back per rep per week, 60% of GTM busywork automated, results that start inside 30 days and mature by 60. The response always ends with the provenance caveat: these are typical results benchmarked against comparable AI-powered GTM implementations and they vary by stack, data quality and adoption, so present them as typical, not guaranteed. Use when a user asks what results to expect. For an estimate on a specific company's numbers use estimate_roi instead. Read-only, static content, no arguments, no authentication, no rate limit.

Takes no parameters.

Returns the phases of a standard 60-day zRev engagement as plain text: what happens in each phase, what is delivered, and when first results appear. Use when a user asks how an engagement works, how long it takes, or what they would receive and when. Read-only, static content, no arguments, no authentication, no rate limit.

Takes no parameters.

Runs zRev's ROI model on one company's numbers and returns, as plain text, the projected annual impact in USD split into pipeline lift, customer acquisition cost savings and the value of hours returned to the team, followed by the model's assumptions and a link to the interactive calculator preset to the same inputs. Use when a user wants a number for their own company; all six inputs are required, so ask for any that are missing rather than guessing. The output is an estimate from a fixed model, not a forecast or a quote. Pure calculation: nothing is stored, no external calls, no authentication, no rate limit.

Fetches https://<domain>/llms.txt (the file that tells AI systems what a site contains) and grades it with seven deterministic checks. Returns a 0 to 100 score and the pass or miss result of each check as plain text. If the site has no llms.txt the response says so and links to zRev's free generator. Use when a user asks whether a site is readable by AI assistants or wants their llms.txt reviewed. Makes one outbound HTTP request to the public domain you pass; stores nothing about it. No authentication. Shares a fair-use limit per IP address with cold_read, and returns a plain-text notice when that limit is reached.

Fetches a company's public homepage and returns, in two or three sentences, what the company appears to do and who it serves, based only on the visible text of that page and with no outside knowledge. Anything missing from the answer is missing from the homepage, which is the point: it shows what an AI assistant would tell a buyer about that company. Use when a user asks how their site, or a competitor's, reads to an AI. The summary is model-generated and can be wrong where the page is vague. Makes one outbound HTTP request to the public domain you pass; stores nothing about it. No authentication. Shares a fair-use limit per IP address with grade_llms_txt, and returns a plain-text notice when that limit is reached.

Returns zRev's go-to-market due diligence checklist for investors and acquirers: 47 questions across ten dimensions, each with the data-room artifact that answers it and the red flag to watch for. Call with no arguments to get the list of ten dimensions; call with a dimension number from 1 to 10, or a keyword such as churn, pipeline or marketing, to get that dimension's questions in full. Use when a user is assessing a company's revenue engine before an investment or acquisition. Read-only reference content, no authentication, no rate limit.

Returns the link and instructions for booking a free 30-minute intro call with zRev's founder, plus what the call covers. It does not book anything: it only returns the scheduling link for the human to open themselves. Use when a user says they want to talk to zRev. If they would rather be contacted, use leave_contact instead. Read-only, static content, no arguments, no authentication, no rate limit.

Takes no parameters.

Records an email address, with optional name, company and note, so that zRev's founder can reach out, and returns a plain-text confirmation. This is the only tool here that writes data: it creates one lead record on zRev's side and triggers no email to the address given. Use only when the human you are assisting has explicitly asked to be contacted by zRev; never call it speculatively or with an address the user has not given you for this purpose. Submitting the same email again updates the existing record rather than creating a duplicate. No authentication. Rate-limited per IP address.