Outbound Offer

Outbound offer (audit + draft)

Draft and audit cold-outreach offers and CTAs only. Read-only analysis and draft text — never send, schedule, or activate outreach. Never invent scarcity, peer proof, or metrics the user did not provide.

Default to the quick shape when the user asks for a draft, rewrite, or "just the offer/CTA." Use the full scorecard only for audits or diagnosis. Render only the selected shape. A single short caveat line is allowed when proof is missing.

For a full-email request, establish the offer and CTA first, then apply the body and subject-line skills when available. Never block if an adjacent skill is unavailable; return this skill's part and state the remaining inputs needed.

Modes

• Audit — score an existing offer/CTA; name failing levers; recommend a rewrite.
• Draft — produce a new offer + soft CTA from inputs (ICP, outcome, proof, ask).

Infer mode from the request. If both are needed, run Audit then Draft.

Core rules (encode verbatim)

Value equation

Value = (Dream Outcome × Perceived Likelihood) / (Time Delay × Effort & Sacrifice).

The formula is conceptual: lower raw delay and effort increase value. For the operational scorecard, use their positive inverses — Speed to Value and Ease — so all four scores use 5 = best. Never divide by the positive scorecard values. Name the weakest lever (usually low Perceived Likelihood or low Ease). Full scoring: [references/value-equation-rubric.md](references/value-equation-rubric.md).

Sell the next micro-commitment, not the product

CTA ladder: interest yes/no → resource offer → 15-min ask → graceful breakup.

Patterns and ladders: [references/cta-patterns.md](references/cta-patterns.md).

Soft CTA defaults

• Soft CTA default (reply ~4.2% vs hard book a demo ~1.4%, ~3×).
• One CTA only (up to ~371% higher CTR).
• CTA
• CTA:
• ·

## Output contract

GitHub-flavored markdown only. Blank line between every block. Every line under a section is a `-` bullet unless noted. Marks: ✓ pass · ✗ fail · · advisory. Lead `### Score / levers` with a fenced `text` meter — one row per lever, bar = `█`×n then `░` to width 5; mark the weakest `✗ weakest`. Bars are a monospace glance only (`█` filled · `░` empty; no colored emoji).

### Shape

````markdown
## Offer 

### Mode
- 

### Score / levers

```text
Dream outcome   █████  5/5
Likelihood      ██░░░  2/5  ✗ weakest
Speed to value  ███░░  3/5
Ease            ████░  4/5
Overall  fair

• Dream Outcome:
• Perceived Likelihood:
• Speed to Value:
• Ease:

Findings

• …

Recommended offer / CTA

• Offer:
• CTA:
• Ladder step:

Why-you / why-now / what-you-get

• Why you:
• Why now:
• What you get by replying:

````

Safety

• Never send, schedule, or activate messages.
• Never fabricate proof, metrics, logos, or scarcity.
• If proof or numbers are missing, mark · and ask for them under Findings — do not invent.
• Keep claims peer-level and specific; strip feature dumps and stacked asks.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: outboundsync
• Source: outboundsync/skills
• License: MIT

Install and usage instructions live in the source repository linked above.