Adversarial 3-round sales review skill for Claude Code. Decide CLOSE, REFRAME, or DISQUALIFY on any deal — pencil to enterprise — with a Buyer-Skeptic persona challenging both your closer and your consultant, judged by Cialdini against the 7 Influence Principles, audited by a discipline foreman.
Status: v1.0.0 — transactional-tier calibrated. Enterprise / strategic-committee tier pending live calibration. See CHANGELOG.md.
You give it a deal (offer + price + buyer profile + last signal). It runs 5 named roles across 3 adversarial rounds and emits one of three verdicts:
- CLOSE — substrate is green, ask for the order on the next contact.
- REFRAME — the offer or position is mis-shaped for this buyer; restructure before next contact.
- DISQUALIFY — wrong buyer, wrong time, or wrong product fit; walk and reinvest the cycle.
Plus a Buyer-Skeptic test — a binary or numeric reality-check the rep can run in 7 days on rep-discretion budget — and a highest-leverage move.
The skill is scale-agnostic. The same 5 roles handle a $2 pencil sale and a $10M strategic-committee deal — weight shifts via the documented scale dial.
| Role | Composite | Owns |
|---|---|---|
| Closer | Belfort + Cardone + Girard | "Transfer certainty, ask for the order now, every no is information." |
| Consultant | Rackham (SPIN) + Ziglar + Holmes + Bettger | "Diagnose pain, earn trust, advance the cycle, get the buyer to articulate cost of inaction." |
| Buyer-Skeptic | Persona challenger | Sophisticated buyer with money + alternatives + scar tissue. Attacks across 3 distinct vectors (price / trust / authority / need / timing / alternatives / risk). |
| Cialdini | 7 Influence Principles + Voss + Kahneman + Aristotle | Judges all 9 prior outputs. Issues CLOSE / REFRAME / DISQUALIFY. |
| Foreman | Discipline auditor | Runs 5 checks: surfaced assumptions, minimum edits, surgical traceability, verifiable test threshold, tunnel-vision dependency map. |
No motivational filler. No "always be closing" without a buyer profile. No "build the relationship" without a forecasted close date.
Three install paths. Pick whichever you prefer.
/plugin install sales-councilVerify: open Claude Code and type /sales-council. The skill description should appear in the slash-command list.
Git does not run install scripts — this step is manual.
macOS / Linux:
git clone https://github.com/ojesusmp/SalesCouncil.git ~/.claude/skills/sales-councilWindows (PowerShell):
git clone https://github.com/ojesusmp/SalesCouncil.git $env:USERPROFILE\.claude\skills\sales-councilVerify:
test -f ~/.claude/skills/sales-council/SKILL.md && echo OKTest-Path "$env:USERPROFILE\.claude\skills\sales-council\SKILL.md"The npm install runs a postinstall script that copies the skill to ~/.claude/skills/sales-council/ automatically.
npm install -g sales-councilOr directly from GitHub:
npm install -g github:ojesusmp/SalesCouncilVerify:
node -e "require('sales-council/install.cjs')" --verifyor simply check the deployed file:
test -f ~/.claude/skills/sales-council/SKILL.md && echo OKTest-Path "$env:USERPROFILE\.claude\skills\sales-council\SKILL.md"Windows + Node 24 + npm 11 note: the postinstall may print
MODULE_NOT_FOUNDand exit with code 1 even though the copy succeeded. This is a known cosmetic quirk on that exact stack — verify with theTest-Pathline above or usenpm pack+npm install -g .\sales-council-<version>.tgzfor a clean exit code.
After install (any path), open Claude Code and type /sales-council. The skill should appear in the slash-command list with its description.
/sales-council <deal context: offer + price + buyer profile + stage + last signal>
sales · sales-council · close-this · sc · sell-it · ask-for-the-order
Pencil sale (transactional, single round):
/sales-council Selling a $2 pencil. Buyer is me right now. I have $5 cash. Last signal: I picked it up.
Mid-market SaaS:
/sales-council $48k/yr CRM seat-expansion deal. Buyer = VP Sales who already uses our product, needs CFO sign-off. Stage: post-demo. Last signal 5 days ago: "send case studies for similar-sized teams."
Enterprise stalled deal:
/sales-council $280k/yr renewal, Fortune-1000 buyer, committee = CFO + Legal + IT procurement. Stage: vendor-risk review, stalled 21 days. Last signal: "send SOC2 + DPA redlines."
- Substrate echo — BANT (Budget, Authority, Need, Timeline) + Stage. Defaults to most-conservative if any field is "unsure."
- Round 1 / 2 / 3 — Closer move + Consultant advance + Buyer-Skeptic objection per round (Skeptic uses 3 distinct vectors).
- Cialdini verdict — deal in one sentence, strongest round per role, verdict (CLOSE / REFRAME / DISQUALIFY), required moves (max 5, each tagged with Influence Principle + Skeptic objection it neutralizes), Buyer-Skeptic test, highest-leverage move.
- Foreman pass — 5 discipline checks, amended final move list, restated test + leverage move.
Token budget target ≤1,700 per run; cap ≤2,000.
- Pure prospecting volume coaching ("how do I 10× my dials") → process problem, not deal problem.
- Compensation plan design.
- Recruiting / hiring sales reps.
- Regulated sales requiring legal review (medical devices, securities, controlled substances) → get counsel, not a council.
- Fraudulent or harmful product → skill refuses at intake.
- Pure ideation with no buyer in mind.
- Marketing asset critique (page, ad, email copy) → use a marketing-council skill.
- Multi-decade brand strategy → out of scope; this is deal-level.
Honest accounting (do not skip):
- Transactional tier (sub-$10k, single buyer): ✅ calibrated 2026-05-13 against a synthetic local-service-business ICP. All 5 internal checks passed (under cap, 3 distinct Skeptic vectors, ≥2 Influence Principles cited, Foreman caught a real economic dependency, verdict actionable).
- Mid-market tier ($10k–$250k, 2–3 stakeholders): ⏳ untested.
- Strategic-committee tier ($250k+, committee + procurement): ⏳ untested.
If you fire this on a strategic-committee deal, treat the verdict as a directional signal, not a binding decision, until the upper-tier calibration ships in v1.1.
Sales Council is one node in a "cabinet" pattern designed for cross-domain decisions:
- Pricing decisions, public promises, scope changes, customer accept/reject — these touch ≥2 of (Sales lane, Marketing lane, Tech lane). Convene a 3-council cabinet, run them in parallel, feed the 3 verdicts into an executive-council arbitrator (CONTINUE / PIVOT / HOLD), audit externally with a Trio skill.
- Single-lane decisions stay with one council.
The cabinet pattern is documented inside the skill but is not a separate file — Sales Council itself is standalone-runnable.
Orlando Molina — TruePoint Agents.
If you fork, adapt, or build on this skill, please keep the attribution in LICENSE and marketplace.json intact so users can trace the source.
MIT — © 2026 Orlando Molina (TruePoint Agents). Fork, adapt, ship.
PRs welcome that:
- Add calibration evidence at mid-market or strategic-committee tier (real or synthetic deals).
- Tighten role definitions without losing distinct adversarial structure.
- Add new Buyer-Skeptic vectors backed by named source.
PRs that will be declined:
- Adding "feel good" language to any role (roles are adversaries, not personas).
- Adding rotating motivational quotes.
- Adding new judges beyond Cialdini (would belong in a different skill).
- Removing the ethical floor (no false scarcity, fake social proof, manufactured urgency, bait-and-switch — non-negotiable).
- Closer: Jordan Belfort (straight-line) · Grant Cardone (10X / pig-headed close) · Joe Girard (Law of 250).
- Consultant: Neil Rackham (SPIN Selling) · Zig Ziglar (value-first) · Chet Holmes (Ultimate Sales Machine / Dream-100) · Frank Bettger (objection handling).
- Judge: Robert Cialdini (Influence + Pre-Suasion 7 Principles) · Chris Voss (Never Split the Difference / tactical empathy) · Daniel Kahneman (System 1/2, loss aversion) · Aristotle (ethos / pathos / logos).
Their work shapes the role behavior; the skill itself is original adversarial-loop architecture, not a remix or derivative product.