Help center.

A PR system for multi-brand operators. One workspace, every brand in your portfolio, modules that share one brief.md per brand and one diligence layer. Four modules are live today: PR-Pitch (HARO, B2B Writer, Substack, X, LinkedIn into one inbox; voice-validated drafts), Editorial Calendar(cross-brand content grid + slot drafts + topic fanouts + recurring weekly slots + auto-publish to Buffer), Outbound (multi-brand cold email + LinkedIn paste-validate + per-brand sequences + unified reply inbox), and Asset Management (versioned brand assets + version-timeline UI + bulk ZIP import + public press-kit page + ZIP). Source Directory ships Q4 2026. All modules included at every tier.

A focused afternoon for the first brand, plus 30 minutes per additional brand if you run more than one. Most of the time goes into refining your voice rules and brand context. The technical setup is fast.

No. The dashboard handles everything. Technical users can use Git and the CLI for power workflows, but they're optional.

Yes. Voice validation, outlet diligence, and submission packaging all deliver value at brand count one. Solo at $49/mo is built for exactly this. If you also run additional brands, multi-brand routing and per-portfolio pricing kick in on Operator and Intel.

Five live sources today: HARO and Help a B2B Writer (email-forwarded); Substack, X (Twitter), and LinkedIn (poll-driven). Reddit infrastructure is built but paused (the relevant subs were banned). If you use a source we have not added, tell us.

Three monthly tiers, same features at every tier. Solo at $49/mo runs 1 brand. Operator at $149/mo runs up to 5 brands. Intel at $249/mo runs up to 10 brands. Every tier gets the full feature set: voice validator on every draft, standalone voice validator, full outlet diligence, API + CLI + MCP access. The bigger plans cost more because they do more work, not because they unlock features.

Per brand: 2,000 scored queries per month and 200 drafts per month. The caps exist to keep a hot account from running our Anthropic bill into the ground. If you genuinely need more, email us. We will lift the cap for legitimate workflows. Caps scale with brand count on Operator and Intel.

A 14-day trial on every plan with no card up front. We don't do an indefinite free tier because the product's value depends on the brief.md foundation, and that requires real authoring time.

Drafts and pitch history stay in your account for export. After cancellation, you download your data and your brief.md files. PRAPI doesn't hold your work hostage.

Your account, backed by a Git repository (we host it by default, or point at your own). Our database is a cache, not the source of truth. If PRAPI disappeared tomorrow, your brief.md would still exist as portable markdown you control. One file per brand. Editable by you, anytime. Exportable. See the brief.md docs hub for the full structure and visibility-scopes model.

Yes. The specification is published. The reference parser will be open-sourced once the spec stabilizes. Anyone can implement it. PRAPI is the canonical implementation, not the only allowed implementation. The Integrity Framework stewards the specification and holds the trademark, structurally separate from PRAPI.

Whoever you decide. Public sections are visible to anyone who fetches your brief URL. Pitch-shareable sections are visible to journalists who receive your pitches. Private sections are visible only to you and people you invite to your workspace. You control all three scopes by tagging sections in your brief.

Your brief is yours. Export it, commit it to your own Git repo, or host it on your own domain. The format is open. Other tools can implement it compatibly. There is no lock-in.

Use PRAPI without publishing your brief publicly. Everything works internally. The protocol benefits compound when you're ready to publish, not before. Most operators don't publish for the first month or two anyway.

No. PRAPI surfaces opportunities, suggests 3-5 angles you could pitch, polishes your own 3-5 sentences of substance into a journalist-ready format in your voice, and packages submissions. You write the substance, you review the polish, you send. The pitch the journalist receives is your words, formatted. Not AI guessing what you would say. The quality discipline is structural; automation without your input produces the kind of spam that ruined HARO.

No. You do. The default flow is reverse-prompt: PRAPI surfaces angles based on your brief.md and what we know about the journalist, you pick one (or write your own), then you type 3-5 sentences of your actual thinking. PRAPI formats that into a pitch, applies your voice rules, attaches submission assets. The substance is yours. There is also a legacy “auto-draft” affordance that writes from scratch without your input; we keep it as a fallback but the assisted path produces better pitches.

It runs on every draft, on every tier. It is structural, not vibes-based. Your banned phrases stay banned. Your one-liner stays consistent. Every tier also gets the standalone validator: paste in any text and check it against your brand voice.

That is the only way to send it. You write, we polish, you send. We never email on your behalf.

Yes. We do not train on your content. Your brief.md, your inbox, your drafts are yours. We use Claude (Anthropic) for drafting and scoring; Anthropic's enterprise API does not train on customer data by default and we are on that contract.

Yes. Settings page exposes a slot for your key. When set, all your scoring + drafting calls bill to your key and bypass our shared platform credit.

Four out of five live: PR-Pitch, Editorial Calendar, Outbound, and Asset Management all shipped early Q2 2026 (ahead of the original Q3 target). Source Directory ships Q4 2026 with a dual-surface UI + REST API + MCP for AI agents. Backlinks targeting is rolling alongside Outbound (V2 shipped, reuses voice validator + outlet diligence + DataforSEO scorer). See the public roadmap for the full module status.

Yes, all three. REST API v1 is live at app.prapi.dev/api/v1 with bearer auth and full MCP-tool parity. The single-binary CLI is at prapi.dev/cli. The MCP server surfaces six tools (list_pitches, get_pitch, draft_pitch, score_query, override_draft, mark_draft_status) to Claude Code, Cursor, and ChatGPT connectors. As new modules ship, REST + MCP + CLI parity ships in the same scope. "API later" isn't an acceptable scope cut.

Backlinks targeting (rolling now; reuses voice validator + outlet diligence + DataforSEO scorer); journalist-brief.md for the two-sided protocol; per-brand outbound podcast and newsletter sourcing; cross-brand analytics. Customer asks shape the order. If you need something fast, email hello@prapi.dev.