Turn your ClaudeCursorCodexHermesClaude into a cofounder who says no.

Yes - the MCP server is live with two transports and the same six tools: five for your own ideas (start_analysis, get_analysis, enrich_analysis, export_analysis, generate_ideas) and get_agency for client work on a Consultant or Agency workspace, which is also what workspace: "agency" on start_analysis, get_analysis and export_analysis acts on. Those Agency calls each need one specific scope - agency:write to start a client analysis, agency:read for get_agency and the two agency reads - and neither is ever granted implicitly, so an older key gets a 403 scope refusal on them until it is replaced or the connector is reconnected. Neither recovery is unconditional: reconnecting only widens the key if the client rebuilds its scope request from our discovery document rather than replaying a cached one, so check the consent screen actually lists the Agency permissions before approving (reconnecting revokes the current key either way), and creating a key by hand needs either a paid personal plan or membership of a Consultant or Agency workspace with an active subscription. It is listed in the Claude Connectors Directory (claude.ai/directory/preuve-ai). The remote server at https://mcp.preuve.ai/mcp works as a claude.ai connector (OAuth sign-in and a consent screen, no key handling) or with any HTTP MCP client sending an Authorization header; the original local stdio server remains fully supported. A ready-made Agent Skill (docs.preuve.ai/agent-skill) teaches your agent the full workflow: scan-type quota safety, polling, enrichment gates, and safe retries. Failed runs refund what they claimed, and error codes are stable so agents can retry safely.

Two ways. Connecting the Preuve connector from Claude signs you in and mints the key for you during consent, on any plan including free. Creating a key by hand in the account panel is self-serve on a paid personal plan or in an active Consultant or Agency workspace. Start free: connect and run your first starter scans. Each key is scoped, sent in one x-preuve-key header, and tied to one Preuve account. A key carries the five personal scopes by default (analysis:write, analysis:read, batch:write, batch:read, export:read). The two Agency scopes, agency:write and agency:read, are never implicit: ask for them when you create the key, or approve them on the OAuth consent screen. They are what unlocks the Agency workspace, and workspace membership is checked separately on every call, so holding them does nothing on an account that belongs to no workspace.

scanType is explicit on every run. "starter" runs the Reality Check at no cost from your Starter scan allowance. "deep" runs the full paid analysis (15+ sections, pivots, citations) and consumes your account quota: subscription scans first, then lifetime monthly scans, then tokens. An agent can never spend paid quota by accident, and a refused deep scan fails loudly instead of silently downgrading.

Polling returns run status, enrichment progress, per-module states, and URLs. The export endpoint returns structured ideas-json (schemaVersion 2): scores, verdict, market, competitors, SWOT, risks, quick take, and citations on every tier, plus the full deep report body (sections, pivots, and generated module payloads) on paid deep runs owned by your key.

Optional add-on generations on paid deep reports, requested through the enrich endpoint: proofOfDemand (verified real-demand prospects), founderFit (founder-vs-plan underwriting from your founder profile), playbook (launch playbook), trends (on-demand Google Trends), and community (community signals from forums, Reddit, review sites, and X). Each module is capped at one successful generation per report and is included in the deep report price - no extra fee.

No. A key is not a login: it cannot open the dashboard, read your account, change billing, or manage other keys. It sees only the analyses it created through the API, and reports you ran in the web app stay invisible to it. The one exception is a key you gave Agency access: agency:read can list and export every client report in your Consultant or Agency workspace, dashboard-created ones included, and agency:write can spend that workspace shared credits. Those two scopes are never implicit, so a key that was not asked for them has no such reach. If a key leaks, revoke it; your account itself is untouched, but revoking an Agency key is urgent in a way a personal one is not.

API usage draws from the same analysis quota as your Preuve plan - Starter scans from your Starter allowance at no cost, deep scans from your subscription, lifetime, or token balance. There is no separate API fee, and failed runs refund whatever they claimed.

Yes. Autonomous build agents such as Polsia execute fast but skip the demand-validation step, which is where most teams burn their runway. Run the idea through the Preuve validation API first, then hand a validated brief to your build agent. See the linked Polsia review for a deeper breakdown of that gap.

Read the Polsia review