agentsclimarketplace

Email auditor

Skill saleshandy/outreach-superpower/skills/email-auditor

Use when user pastes existing email copy and wants it reviewed, scored, or improved. Triggers on "audit my emails," "audit this email," "review this copy," "review my cold email," "score this email," "rate my outreach," "rate this copy," "is this email good," or any pasted email body with subject + body structure.From its SKILL.md

Install
npx -y skills add saleshandy/outreach-superpower --skill email-auditor

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

8.7 KB, as published. Nobody here has run it

Email Auditor

Audit pasted cold email copy across 11 dimensions (7 structural + 4 craft), rate each 1-5, list deliverability red flags from references/inbox-radar.md, output strengths + rewritten suggestions, and persist the audit to a dated file.

Inputs

  • Required: pasted email copy (detect subject + body)
  • Optional (helpful): icp.md in workspace - enables context-aware critique

Workspace path

Default campaign: default. If user invocation includes --campaign <name>, use that. Full rule: see using-outreach-superpower.

Process

Step 1 - Detect copy structure

Parse the pasted text for:

  • Subject line: look for explicit Subject: prefix first. If absent, fall back to: first non-empty line that does NOT start with greeting tokens (Hi, Hello, Hey, Dear, Good) AND is shorter than 80 characters.
  • Body: everything after the subject line.
  • Sender name / signature (optional): bottom-of-paste lines after a blank line.

If no subject line is detectable by either rule, ask once: "What's the subject line?" Then re-run detection with the supplied subject before proceeding to Step 2.

Step 2 - Read context (if available)

Read outreach-workspace/<campaign>/icp.md if it exists. Extract:

  • Primary segment name (e.g., "VP Engineering at Series A SaaS")
  • Buying motion (self-serve / demo / both - from icp.md fields)
  • Top 1-2 angles from the Angle Kit (if present)

Set icp_aware: true in the audit frontmatter. Use the extracted info during Step 3 scoring (see ICP-aware adjustments below).

Step 3 - Score 11 criteria (1=Poor, 5=Excellent)

3a. Structural dimensions (7)

CriterionWhat to check
Subject LineSpecific enough to convey topic; <60 chars; personalization present when ICP supports it; curiosity without clickbait; no spam triggers
Opening LineGrabs attention; personalized; relevant to recipient's industry/role; avoids "Hope you're doing well"
Body CopyClear, concise, valuable; recipient-focused (not product-focused); credibility/social proof present; <150-200 words
Call-to-ActionClear next step; encourages reply; binary CTA preferred ("Open to a 10-min call?")
Formatting & ReadabilityShort sentences; no walls of text; conversational tone; mobile-friendly
Personalization & Human TouchReferences work/company/industry; feels 1:1 not mass; empathetic
Spam & DeliverabilityNo spam triggers, excessive caps, exclamation marks; <=1 link; subject <60 chars; body <200 words

3b. Craft dimensions (4) - from PB-02

CriterionWhat to check
Tone matchCopy matches the ICP brand voice / segment expectations (professional / conversational / witty / bold / warm). If icp.md is absent, infer tone from the copy itself and call out internal inconsistency (e.g., warm opener, aggressive CTA). Bands: 1 = register clashes with ICP (bro-y for enterprise, stiff for creator/SMB); 3 = consistent register but generic, not tuned to segment; 5 = register clearly matches ICP brand voice and stays consistent across opener/body/CTA.
Hook strengthFirst-line specificity, relevance, curiosity factor. Bands: 1 = generic opener ("Hope this finds you well," "I came across your company"); 3 = relevant but unspecific (role/industry reference without named context); 5 = specifically references the recipient's measurable context (named achievement, named pain, named recent event).
Spintax usageCount of {spin}...{endspin} blocks; presence in high-frequency phrases (greeting, body, sign-off, CTA). Bands: 1 = zero spin blocks (no variant diversity, fails A/Z testing); 3 = 1-2 blocks present but missing from high-frequency phrases; 5 = >=3 blocks covering greeting, body, and sign-off/CTA, with tonally consistent variants.
Word economyBody word count. Target <120 for email body (PB-02: 60-100 for opening, 50-70 for follow-up). Bands: 1 = body >150 words with preamble and filler sentences (>25 words each); 3 = within range but contains 1-2 lines that fail the "So What?" test; 5 = within range, every sentence earns its place, no preamble.

ICP-aware adjustments (only if icp.md was loaded):

  • Personalization & Human Touch: weight against segment fit. Flag if email targets generic "founders" when ICP says "VP Engineering at Series A SaaS."
  • Call-to-Action: flag mismatches with the buying motion (e.g., "Book a 30-min call" CTA when ICP says self-serve preferred).
  • Body Copy: flag if value prop doesn't match the segment's primary pain.
  • Tone match: weight against the ICP's declared brand voice or segment expectations. State the segment name explicitly when calling out a mismatch.

State the segment name explicitly in feedback when raising an ICP-related issue: "Doesn't fit your ICP segment 'VP Engineering at Series A SaaS' - they prefer self-serve."

Step 3c - Deliverability red flags

After scoring the 11 dimensions, run a focused deliverability scan against references/inbox-radar.md. This is a separate check (not double-counted in the Spam & Deliverability score) that surfaces concrete risk patterns the user should fix before sending.

Scan for:

Red flagSource rule
Spam-trigger words (e.g., free, guarantee, click here, act now, 100%, urgent, winner)inbox-radar.md "Content suggestions" + PB-02 spam-word list
Body word count outside 60-150 word ideal rangeinbox-radar.md "Content suggestions"
Subject >60 characters or ALL CAPS or 3+ exclamation marksinbox-radar formatting concerns
Multiple links in opening email (>1 link, or any link in step 1)PB-02 + inbox-radar
Imbalanced text-to-link or text-to-image ratio (e.g., 1 link in 5-word body)inbox-radar
Generic unverified merge tags without fallbacks ({{First Name}} without |'there')PB-02 merge-tag rules
Subject line uses urgency stack: "URGENT", "ACT NOW", "Limited time"inbox-radar spam-trigger words

Output format for this section in the audit:

### Deliverability red flags

- **Spam trigger detected:** "free trial" in body line 3. Suggest: "no-commitment pilot" or "14-day test run."
- **Length warning:** body is 187 words; ideal range is 60-150 (per `references/inbox-radar.md`).
- (...if no flags found, write: "No red flags detected. See `references/inbox-radar.md` for the full deliverability framework.")

If the email has 3+ red flags, recommend the user run an Inbox Radar test before sending and cite the reference.

Step 4 - Output to chat

## Audit: <subject line preview>

**Overall: X/55** (sum of 11 criterion ratings)
**Structural: X/35** | **Craft: X/20**

| Criterion | Rating | Feedback |
|---|---|---|
| Subject Line | X/5 | ... |
| Opening Line | X/5 | ... |
| Body Copy | X/5 | ... |
| Call-to-Action | X/5 | ... |
| Formatting & Readability | X/5 | ... |
| Personalization & Human Touch | X/5 | ... |
| Spam & Deliverability | X/5 | ... |
| Tone match | X/5 | ... |
| Hook strength | X/5 | ... |
| Spintax usage | X/5 | ... |
| Word economy | X/5 | ... |
| **Overall** | **X/55** | **Structural X/35 + Craft X/20** |

### Deliverability red flags
- (list from Step 3c, or "No red flags detected" if clean)

### Strengths
- ...

### Improvements
- **Subject:** <rewritten suggestion>
- **Opener:** <rewritten suggestion>
- **Body:** <rewritten suggestion>
- **CTA:** <rewritten suggestion>
- **Spintax:** <added variants if missing>

*(Improvements list pointed fixes per component; Rewritten draft is the synthesized paste-ready copy.)*

### Rewritten draft
Subject: ...
Body: ...

Step 5 - Write audit file

Filename: audits/YYYY-MM-DD-<slugified-subject>.md. Slug = first 5 words of subject, lowercase, hyphenated. If file exists, append -2, -3, etc.

Frontmatter:

---
audited_at: YYYY-MM-DD
overall_score: X/55
structural_score: X/35
craft_score: X/20
deliverability_flags: <count>
icp_aware: true | false
subject_preview: <first 60 chars>
---

Body: include the original copy (for reference), then the full audit output above (11 dimensions + deliverability red flags + strengths + improvements + rewritten draft).

Step 6 - Cite the source

  • If icp.md was used, end the audit with a one-liner: "Critique informed by icp.md segment '<segment name>'."
  • If any deliverability red flags fired, end with: "Deliverability framework: references/inbox-radar.md. Re-run an Inbox Radar test after revisions."

Keep looking

Skills are one crate of 628,853. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.