Run an AI SDR play on new leads
AI SDR agent is an Agent Skill: a written method your AI applies on demand, in Zealos or in any client that reads SKILL.md.
AI SDR agent is an Agent Skill: a written method your AI applies on demand, in Zealos or in any client that reads SKILL.md.
An AI SDR agent is only as good as the definition of a good customer it is working from, which is why the fit criteria belong in a written skill rather than buried in a prompt nobody can find.
| name | our-icp |
|---|---|
| description | Who we sell to, the signals that indicate fit, and the disqualifiers. Use when scoring a lead, deciding whether to pursue an account, building a prospect list, or writing targeting copy. |
| allowed-tools | get_icp_context score_lead_against_icp research_account search_companies |
| compatibility | Zealos, Claude Code, Cursor, Codex and any client that reads SKILL.md |
| license | Free to use and edit |
--- name: our-icp description: Who we sell to, the signals that indicate fit, and the disqualifiers. Use when scoring a lead, deciding whether to pursue an account, building a prospect list, or writing targeting copy. --- # Our ICP Start by calling `get_icp_context` — the workspace's configured profile is the source of truth, and this skill is how to *reason* about it, not a copy of it. ## How to score Weight in this order. Later signals cannot rescue a failure on an earlier one. 1. **Pain we actually solve.** Do they have the problem, today, at a size that hurts? Not "could benefit from". 2. **Trigger.** Something changed recently: new leader in the function, funding, a public commitment, a competitor switch, a hiring spike in the team we serve. Without a trigger, a good-fit account is still a cold one. 3. **Reachability.** Can we get to the person who feels the pain, not just to someone adjacent? 4. **Firmographics.** Size, industry, geography, stack. Last, not first — firmographics are a filter, not a reason. ## Disqualifiers These end the evaluation regardless of anything else: - No budget owner in the function we serve. - Bought a direct competitor in the last 6 months. - The pain is real but owned by a team we cannot reach. - Requires a feature we do not have and are not building. ## Output A score, the two strongest signals with evidence, the strongest counter-signal, and the recommended next action. If the counter-signal is a disqualifier, say "disqualified" plainly — do not average it away against strong firmographics.
The list you work is already filtered by the rules your team actually uses, so the fifty accounts in front of you are fifty you would have picked yourself.
Rep-to-rep variation in what counts as qualified collapses, because the criteria live in one document instead of in six people's heads.
Your first hire inherits your definition of a good customer rather than reconstructing it from the deals that happened to close.
An agent that runs the repeatable part of sales development — research, list building, first drafts — on a trigger or schedule. The skill it applies is what makes its output specific to your company.
Only if you enable that tool. Most teams start with drafts in a review queue and turn on sending once the output stops needing edits.
A model gives you a number you cannot argue with. A skill gives you criteria you can read, disagree with and edit — which matters when the definition of a good customer changes every two quarters.
Post-call automation turns a recorded conversation into the three things that outlive it: the commitments each side made, the fields that changed, and the follow-up that has to go out today.
/call-recap-formatCRM data hygiene survives AI only if one rule is written down first: a field left blank is a known unknown, while a field filled in by inference is an unknown unknown that will be reported on as fact.
/crm-field-conventionsAccount research earns its time only when it changes the first sentence you write, which means looking for what recently changed rather than assembling a profile anyone could have found.
/our-icpQualification is a claim about evidence, not a score: a deal is qualified when every criterion has a source you can quote, and unevidenced criteria are named rather than filled in optimistically.
/meddiccA follow-up gets a reply when it opens with the specific commitment made on the call, stays under 120 words, and asks for exactly one thing the recipient can answer in a sentence.
/email-rulesA pipeline review is useful in proportion to how uncomfortable it is: the job is to find the deals that are slipping quietly, not to narrate the ones everybody already knows about.
/meddicc