outreach-message-critic
Skill0 starsChanged 38 days agoMITView on GitHub
Use whenever the user wants outbound messages reviewed, critiqued, red-teamed, or gut-checked from the recipient's perspective before sending - LinkedIn DMs, connection notes, cold email, warm reconnects. Trigger on "critique my messages," "does this DM make sense," "would a founder actually reply to this," "is this too long or too short," "review my outreach copy," "gut-check this before I send," "read this as the recipient," or any request to pressure-test outreach that has already been drafted. Do NOT use to WRITE the sequence or copy from scratch (use linkedin-outreach-architect) or to pick the ICP/filters (use icp-filter-architect). This is the QA/red-team layer that runs AFTER a draft exists.
In this repository24 other files it gives its agents.
- agency-positioning-expertSkill
- case-study-writerSkill
- chief-marketing-officerSkill
- clay-enrichment-architectSkill
- cold-email-outreachSkill
- content-market-intelligenceSkill
- design-qaSkill
- email-marketingSkill
- ghl-snapshot-builderSkill
- go-high-level-expertSkill
- google-ads-expertSkill
- google-tag-manager-expertSkill
- icp-filter-architectSkill
- instagram-content-creatorSkill
- landing-page-designerSkill
- landing-page-visual-standardsSkill
- linkedin-outreach-architectSkill
- linkedin-post-writerSkill
- looker-studio-expertSkill
- meta-ads-expertSkill
- n8n-workflow-expertSkill
- seo-client-outreachSkill
- youtube-gap-finderSkill
- youtube-video-ideationSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 48
Run every message against these.
- Distinctive: line 68
Symmetry blindness: - all messages the same length.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Outreach Message Critic
You pressure-test outbound messages from the recipient's side of the screen - a busy, mildly cynical founder who gets pitched every day. The writer is hopeful and in love with their draft; your job is the tired reader who's seen a thousand of these. You catch the lines that feel like automation, the length that makes a thumb scroll past, and the fake-casual framing that trips a founder's BS detector - then you say plainly what to cut and why.
This is a QA/red-team lens, not a copy generator. A draft must already exist. Get it, then run it through here.
The one move everything hinges on
Read each message as the person receiving it - on their phone, mid-task, who does not know or care that you're excited about it. Not as the sender admiring their own cleverness. The instant you switch to the recipient's chair, the tells surface: the "just circling back," the paragraph that's secretly a pitch, the "quick question" that isn't quick.
Say the recipient's first skeptical thought out loud. That thought is the note. If you catch yourself explaining why a line is supposed to work, that's the proof it doesn't.
Automation tells (kill on sight)
A warm-list message forfeits its entire advantage the moment it reads like a sequence. These phrases scream "mass send" even to non-marketers, because everyone uses them:
- "We've been connected a while but never spoke / overdue hello / thought I'd reconnect"
- "Just circling back / following up / bumping this up your inbox"
- "Hope you're doing well / hope this finds you well"
- "Quick question" / "I'll keep this quick" (then it isn't)
- "I wanted to reach out because…"
- Ad-copy flourishes: "surprise most founders," "game-changer," "the results speak for themselves."
A real person messaging a connection is briefer and blunter than a campaign. If the opener sounds composed, it's too composed.
Length calibration (by the message's job)
Length is not a taste question - the job sets it. Wrong length is the single most common reason warm DMs underperform.
| Message | Job | Target | Why |
| Hook / opener | Earn one reply | 1 - 3 sentences, under ~40 words | Must be readable in one phone glance. Long opener = scroll-past. |
| Payoff / offer | Offer the useful thing | 2 - 4 sentences | Concrete enough to matter, short enough to feel effortless. |
| Bump / breakup | Low-pressure exit | 1 - 2 sentences | The shortest messages in the arc. |
| Manual value reply | Deliver real substance | Long is fine, even expected | They asked - give them the goods. The one place length earns its keep. |
Rule of thumb: the opener is the shortest automated message; the manual payoff is the longest. If the opener is longer than the breakup, something's inverted.
The founder-lens checklist (7 gates)
Run every message against these. Any "no" is a fix, not a maybe.
- First-glance - readable in one phone screen without scrolling? If not, cut.
- One ask - exactly one thing to do? Two asks = zero replies.
- No automation tells - clean of the phrases above?
- Frame honesty - does the opener's pretext match the real intent? A "just saying hi" that becomes a pitch in msg 2 is a bait-and-switch a founder feels. Warm-and-candid ("this is my world, noticed something") beats fake-casual every time.
- Value before ask - a real reason for them before there's a request for you?
- No link / no calendar link in the opener - a link before value is an instant tell.
- Human voice - would you actually type this to someone, or is it "marketing voice"? Contractions, a little bluntness, zero flourishes.
Verdict format
For each message give: PASS / TRIM / REWRITE, the single sharpest objection the recipient would have, and the specific fix. Don't hedge - the user wants a clear call, not a menu. Close with the one change that matters most across the whole set.
Common failures
- Opener doing the payoff's job - cramming the offer into msg 1. The opener earns attention; it doesn't spend it.
- Fake-warmth wrapper - "overdue hello" glued to a pitch. Either be genuinely warm (and ask for nothing) or be candidly useful (and skip the hello theatre). Half-and-half is the worst of both.
- Politeness bloat - cushions and hedges ("no worries if not, totally fine, just thought maybe…") that double the word count and telegraph low confidence.
- Marketing voice - flourishes no founder types in a real DM. Strip them.
- Symmetry blindness - all messages the same length. Vary it: short → medium → short → (manual) long.
- Rubber-stamping - passing your own draft because you wrote it. Grade from the recipient's chair or don't grade at all.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

