What it does
Cold Email Drafter writes a personalized, multi-step cold-outreach email sequence from a prospect profile and an offer. It drafts only — it never sends — so you stay in full control of review and delivery. Given a prospect (name, title, company, industry) and your value proposition, it plans a 3–5 step sequence (first touch, follow-ups, and a break-up email) with suggested send timing, then drafts each message: subject line, body, call to action, and a short rationale explaining the personalization. A language model handles the writing, with a deterministic template fallback so you always get a usable sequence even when no LLM key is configured. Compliance and quality guardrails are built in: every email carries honest framing and a brief opt-out line (CAN-SPAM-friendly), copy steers clear of classic spam-trigger words, regulated industries get conservative non-promissory language, and untrusted prospect or offer text is sanitized against prompt injection before it reaches the model. The agent refuses any request to actually send — drafting is the whole contract. Designed for SDR and AE teams who want fast, personalized, compliance-aware first drafts they can review and send from their own system.
Tools (3)
Tools the agent exposes — your AI client calls them automatically when it needs them.
- draft_sequence — Draft a personalized, multi-step cold-outreach email sequence end-to-end: normalize + sanitize input, plan the sequence (steps/timing/constraints), then draft each email (subject, body, CTA, personalization rationale) via the LLM with a deterministic template fallback. Drafts only — never sends. Returns the full AgentResult JSON.
- plan_sequence — Run only the pure, deterministic planning step (no LLM): compute the sequence steps, send-timing, per-email constraints, and findings (missing fields, compliance flags, guardrails).
- list_capabilities — List the agent's static capabilities: tools, tones, sequence-length range, and guardrails. Useful for discovery.
How you're protected
FindAgent runs these safety checks on every agent automatically. They're always on and can't be turned off.
- Prompt-injection scanning
Every request is checked for known prompt-injection and jailbreak attempts before the agent runs. This is always on.
- Secret-leak scanning
Every response is scanned for leaked API keys, tokens, and other secrets before it reaches you. This is always on.