# Lead Scorer — AI SDR that finds, scores and drafts to B2B leads, with the whole workflow exposed as an MCP server. > Source: The Agents Index — https://theagentsindex.com/lead-scorer (structured, researched, re-verified) > Facts last verified: 2026-09-18 Lead Scorer is an AI SDR that discovers B2B companies on the web, scores the account and the decision-maker, and drafts LinkedIn and email sequences. The same product is exposed as an MCP server other assistants can drive. | Fact | Value | | --- | --- | | Website | https://lead-scorer.com | | Price | from €99/month billed monthly, €82.50/month billed annually | | Free tier | Yes | Free — LinkedIn CRM at €0/month: full CRM (leads, companies, lists, segments, tags), LinkedIn CSV and Chrome extension import, Pipedrive-style pipeline, CSV export and MCP server access. The comparison table shows no monthly credits and no AI budget on Free, and web discovery, AI enrichment, two-level scoring, signals, the outbound SDR agent and connected email/LinkedIn sending are all absent from the Free column. ## What it does **Done-for-you service** · Partly — Mostly self-serve software, but the top tier adds real human work: a setup interview, hands-on implementation of the acquisition strategy and a monthly review, on top of the same Pro product capacities. > We implement the acquisition strategy for you > — https://lead-scorer.com/#pricing · Pricing section, Done for You plan card, second bullet of 'Everything in Pro, plus the service' ## How much it does unattended **Runs autonomously** · Partly — Discovery, enrichment, scoring and drafting run end to end without a human, but the send is a hard stop: the vendor states there is no auto-send switch and launching is a human action in the app. > Nothing leaves your account until you approve it. You set the daily cap and the sending window, and every run replays step by step so you can see what the agent did and why. > — https://lead-scorer.com/ · Proof section, card headed There is no auto-send switch **Multi-agent** · Yes — A fixed vendor pipeline rather than a crew you compose: separate agents handle discovery, enrichment and scoring, a drafting agent writes, and a second model proofreads every message before review. > A second model (Mistral) proofreads and optimizes every draft before it reaches your queue. > — https://lead-scorer.com/llms.txt · Section 'Writes the messages — and a second AI reviews them', third bullet **Agent permissions** · Yes — Permissions are explicit and bounded: an agent can build an entire campaign and write every draft, but cannot launch it, and spending is gated by an estimate, a confirmation threshold and a per-call lead ceiling. > Your agent can build the whole campaign and still not send it. Launching is a human action in the app, under your daily cap and sending window. > — https://lead-scorer.com/agent · Guardrails you inherit, Approval before send card ## Which models it runs on **Claude** · Yes — Claude is a first-class client through the hosted MCP server, including Claude Code, with per-client setup guides and OAuth sign-in rather than a pasted key. > Claude (custom connectors), Claude Code, ChatGPT (developer mode), Codex, Cursor, or any MCP client you write yourself. > — https://lead-scorer.com/docs/mcp · Supported clients section **GPT** · Yes — ChatGPT in developer mode and Codex are supported MCP clients, and the vendor markets driving the product from ChatGPT on the same footing as Claude. > Connect Lead Scorer to Claude or ChatGPT and let your AI pilot your outreach and your content creation. > — https://lead-scorer.com/ · MCP built in section, introductory paragraph **Open models** · Partly — Any MCP client you write yourself can drive the product, so the driving model is yours to choose; the vendor never names an open-weight model and its own hosted generation is not selectable. > Works with Claude (custom connectors), Claude Code, ChatGPT (developer mode), Codex, Cursor, or any MCP client you write yourself. > — https://lead-scorer.com/llms.txt · Section 'MCP — bring your own agent', client bullet **Model choice** · Partly — Inside the app the vendor picks the models and bills them against an AI budget; over MCP you supply the model, because classification and drafting happen in your assistant and Lead Scorer only stores the result. > When your agent classifies a company or writes a first touch, the inference already happened — in your model, on your tokens. > — https://lead-scorer.com/agent · Bring your own model, keep your credits section ## Where you use it **In your editor** · Partly — No purpose-built editor extension; Cursor reaches the product as a supported MCP client, and Cursor specifically needs a Bearer API key because it does not yet support the OAuth flow. > Claude (custom connectors), Claude Code, ChatGPT (developer mode), Codex, Cursor, or any MCP client you write yourself. > — https://lead-scorer.com/docs/mcp · Supported clients section **On the command line** · Partly — There is no Lead Scorer CLI of its own; terminal use means connecting the MCP server from Claude Code or Codex, for which the vendor publishes a plugin bundling the OAuth MCP and its skills. > Universal Codex/ChatGPT and Claude Code plugin source: `.agents/plugins/lead-scorer-outreach` in the Lead Scorer repository. It bundles the OAuth MCP and the same generated Skills. > — https://lead-scorer.com/llms.txt · Agent skills section, plugin bullet **In a browser** · Yes — The hosted web app is the primary surface and the only place a campaign can be launched: you brief, review runs and approve drafts there. > Sign up, brief the agent in plain language, review runs and approve drafts in the interface. The CRM is free forever, with no credit card and no trial clock. > — https://lead-scorer.com/docs/getting-started · Two ways to drive it, The app column ## Whose machine it runs on **Self-hosted** · No — Every documented surface is vendor-run: the web app, the Chrome extension and one hosted MCP endpoint. The Terms grant only a limited right to use the service, and no plan offers a customer-hosted deployment. > Subject to these terms and your plan, you receive a limited, non-exclusive right to use the service. > — https://lead-scorer.com/terms · Section 7, Intellectual property **Open source** · No — The product itself is proprietary; what is published openly is the playbook layer, 25 downloadable SKILL.md files, not the application the vendor retains rights in. > We and our licensors retain rights in Lead Scorer's software, interfaces, trademarks and documentation. > — https://lead-scorer.com/terms · Section 7, Intellectual property, first sentence ## What it costs to run **How it meters** · Yes — Two meters on top of a monthly subscription: enrichment and scoring consume credits with a published per-action price list, while drafting draws on a fixed monthly AI budget the vendor says avoids overage. > Two meters, deliberately separate. Credits pay for enrichment and scoring — the work that hits external data sources. > — https://lead-scorer.com/docs/core-concepts · Credits and AI budget section **Free tier** · Yes — A standing free plan, not a trial: the full CRM, network import, pipeline, CSV export and MCP access with no card and no clock, though AI enrichment and the SDR need a paid plan or credit pack. > No credit card, no trial clock — free forever > — https://lead-scorer.com/#pricing · Pricing section, Free — LinkedIn CRM card, final bullet **API access** · Partly — Programmatic access exists but only as the MCP server, authenticated by OAuth 2.1 or a Bearer MCP API key; no public REST API or endpoint reference is documented anywhere on the site. > An Authorization: Bearer header with an MCP API key works as a manual fallback for clients without OAuth support. > — https://lead-scorer.com/docs/mcp · Endpoint and authentication section, second bullet **MCP server** · Yes — MCP is the architecture rather than an add-on: the vendor states every feature ships as a tool, included on every plan including the free one, with the same credits and the same approval gate. > Lead Scorer was built AI-first: every feature ships as an MCP tool rather than as a bolted-on integration. Anything you can do in the app, an agent you control can do over the protocol > — https://lead-scorer.com/docs/mcp · MCP server page, opening paragraph **Bring your own key** · Partly — You can bring your own model but not your own keys to the app: over MCP your assistant does the scoring and drafting on your tokens, while in-app AI runs on the vendor's providers and budget. > When your agent scores a lead or writes a draft, the inference happened in your model, on your tokens, and Lead Scorer only stores the result. > — https://lead-scorer.com/agent · Questions section, answer to 'Why are most of the tools free in credits?' ## Buying it for a team **A company can buy it** · Partly — An organisation can contract for it and billing details include organisation data, but every published plan is a single-user workspace with one connected inbox; no team or enterprise plan is offered. > You must have legal capacity to enter into a contract and, when acting for an organisation, authority to bind it. > — https://lead-scorer.com/terms · Section 1, Provider and scope **Seat model** · No — Not sold per seat at all: a plan buys one workspace metered by connected channels and credits, so scaling means more credits or channels rather than more users. > 1 email + 1 LinkedIn · 1,000 credits · €40 AI included > — https://lead-scorer.com/#pricing · Pricing section, Pro plan card subtitle **Pooled budget** · Yes — There are no seats to split: credits and AI budget sit at account level and are drawn down identically whether the work is triggered in the app, by the agent or over MCP. > The same action costs the same wherever you trigger it — the app, the agent and the MCP server all bill through one path. > — https://lead-scorer.com/docs/core-concepts · Credits and AI budget section, first bullet **Admin controls** · Partly — Real constraints exist on what the agent may do — spend estimates, a confirmation threshold, a per-call lead ceiling, daily caps and sending windows — but they are account-owner settings, with no admin console or roles published. > The three billable tools estimate before they run, require an explicit confirmation above the threshold, and refuse any call targeting more than 250 leads. > — https://lead-scorer.com/agent · Guardrails you inherit, Spend guardrail card **Audit log** · Partly — Runs replay step by step and memory revisions are immutable, and the vendor confirms server-side records of API and MCP activity, but no admin-facing audit log or export is documented. > Server-side operation records include API/MCP activity and AI diagnostic traces. They are distinct from browser cookies; disabling browser tracking does not delete those records. > — https://lead-scorer.com/privacy · Section 7, Cookies, local storage and analytics **Single sign-on** · Partly — Sign-in is email or Google identity, and agents authenticate to the MCP server with OAuth 2.1; no SAML or enterprise OIDC federation is offered or mentioned anywhere on the site. > Google sign-in supplies the identity information you authorise. > — https://lead-scorer.com/privacy · Section 2, Data and sources, Account and billing bullet ## What happens to your code **Data residency** · Partly — The operator is a French SAS under GDPR and the CNIL, but the vendor explicitly warns that processing may leave the EEA for the United States and offers no region choice. > Providers and connected services may process data outside the European Economic Area, including in the United States. European hosting of one component does not mean all processing stays in Europe. > — https://lead-scorer.com/privacy · Section 5, International processing **Getting out** · Yes — CSV export of leads and companies is a free-plan feature and the Terms point to export before closure; what is not stated is whether monthly credits roll over or expire. > Export the data you need using the available export features, or contact us for assistance. > — https://lead-scorer.com/terms · Section 9, Suspension and termination **Certifications** · Partly — GDPR posture is documented in detail, including an Article 28 processing agreement on request and a named subprocessor annex, but no SOC 2 or ISO 27001 certification is claimed anywhere. > That processing requires an agreement meeting Article 28 GDPR, covering the scope, security, subprocessors, rights requests, incidents, transfers and return or deletion of data. > — https://lead-scorer.com/terms · Section 6, Data and confidentiality _Not established: Opt out of training_ ## Provenance Every fact above comes from the vendor. Nothing here is independently corroborated yet. ## Coverage elsewhere Pages about this product that the vendor does not control. The label states how independent each one is. | Source | Kind | Read | | --- | --- | --- | | recherche-entreprises.api.gouv.fr | Independent source | 2026-08-26 | | trustmrr.com | Independent source | 2026-08-26 | ## What changed - 2026-09-18 — Capabilities · Commercial terms · Free tier - 2026-09-02 — Capabilities · Commercial terms · Free tier · Price - 2026-08-31 — Capabilities · Commercial terms · Free tier · Price - 2026-08-26 — Alternatives · Documentation link · LinkedIn link · One-liner · Summary ## Correcting this record Anything marked "Not established" is a question we have not answered — it is not a No. If you represent Lead Scorer, or anything here is wrong or out of date, you can supply it with a source: the operations this site accepts from an agent are listed at https://theagentsindex.com/.well-known/agents.json — that catalogue needs no key to read, and each entry states its own auth. Supplying evidence is free. Nothing that can be bought here changes what a field says, whether a gap is closed, or how this record is scored.