Run any Skill in Manus with one click

$pwd:

b2c-marketing-voice

Name: B2c Marketing Voice
Author: elophanto

// Anti-slop principles for B2C social posts and outreach. Use BEFORE drafting an email, X post, or prospect message so the body reads like a person, not an LLM. Pairs with the Phase 10 voice contract (data/companies/<slug>/voice.yaml) — this skill is the meta-principles; voice.yaml is the per-company specifics. Distilled from public marketing patterns (jackfriks/b2c-marketing on ClawHub) — not a copy.

Run Skill in Manus

$ git log --oneline --stat

stars:88

forks:10

updated:May 26, 2026 at 10:31

SKILL.md

readonly

name	b2c-marketing-voice
description	Anti-slop principles for B2C social posts and outreach. Use BEFORE drafting an email, X post, or prospect message so the body reads like a person, not an LLM. Pairs with the Phase 10 voice contract (data/companies/<slug>/voice.yaml) — this skill is the meta-principles; voice.yaml is the per-company specifics. Distilled from public marketing patterns (jackfriks/b2c-marketing on ClawHub) — not a copy.
license	MIT
metadata	{"author":"petr-royce","version":"1.0.0","phase":10}

B2C marketing voice — anti-slop rules

When you're drafting an outbound message (email, prospect outreach, X post, LinkedIn) for a B2C audience, the default LLM voice is the failure mode. Read this skill, then write. Then run voice_lint to catch what slipped through.

Triggers

email_draft
outreach_draft
post_draft
twitter_post
write a tweet
write a post
cold email
outreach message
social media post

The four dead patterns (do NOT use)

Self-focused hook. We help businesses leverage AI to... / Our team is excited to... / At <Company> we believe... — reader scrolls past in 0.3s. Open on THEM or on a CONCRETE OBJECT.
Generic CTA. DM me, check the link in bio, book a call, learn more — these are not CTAs, they are noise. Either soft-CTA (if this resonates...) or zero-CTA (let the post stand).
Corporate filler. leverage, unlock, synergy, in today's fast-paced world, are you tired of, imagine if, let me explain why. These are LLM tells. Any one of them and the reader knows.
Abstract claim with no object. AI is changing everything / productivity is the new luxury. No screenshot, no number, no person — nothing for the reader to anchor on.

The four winning shapes (use one)

Another person + conflict + showed them + changed mind. My dad didn't believe agents could ship real code. I showed him the PR diff. Now he asks me daily what they shipped.
POV: . POV: you check your agent's overnight log and find 4 closed tickets you didn't write.
I used to . Then . Now . I used to think prompts were the moat. Then I watched a 12B model out-plan GPT-4 on tool use with the same prompt. Now I think the orchestration is.
Concrete object up front. A screenshot, a number, a single observed line of code, a single quoted sentence from a user. Then your one-line take. Then stop.

Length defaults (override per voice.yaml)

X post: 80-240 chars. If you wrote 280 you padded. Cut.
Cold email body: 40-80 words. Subject is its own contract — no preamble in body.
LinkedIn post: 400-800 chars. The hook line stands alone; the rest is one block.

Before you draft

voice_show — if has_voice=True, the company's voice.yaml is the final authority and overrides anything in this skill.
Pick one winning shape from the list above. Name the shape in your reasoning so the operator can audit your choice.
Identify the CONCRETE OBJECT or CONCRETE PERSON. If there isn't one, don't draft yet — go find one (call browser_navigate, ask the operator, read a real prospect's bio).

After you draft

voice_lint(text=<body>) — fix any violations before calling the draft tool. The draft tool will lint again; catching it here saves a round-trip.
Reread the opening line out loud. If it sounds like a press release, rewrite it.
Reread the CTA. If it's DM me or check the link, delete it.

What this skill does NOT do

It does NOT replace the per-company voice.yaml. voice.yaml wins on every conflict.
It does NOT apply to B2B technical writing (changelog, docs, internal memos) — those have different conventions.
It does NOT produce hooks for you. It tells you which shapes work and which fail; the operator's exemplars (and your judgement) produce the actual line.

related-skills.json

same repository

drive-business.md

from "elophanto/EloPhanto"

Canonical workflow for "drive my business" / "operate this company". Single entry point — branches internally to onboard (new), Phase 11 plan (existing without strategy), or execute (existing with strategy).

2026-05-2788

strategy-pipeline.md

from "elophanto/EloPhanto"

Phase 11 procedure — capabilities → set_strategy_inputs → plan → apply → approve, plus 3-path blocker resolution (ask / build / defer). Covers WHOLE business surface via injected OPERATIONAL CONTEXT, not just what_we_sell.

2026-05-2788

trust-ladder-workflow.md

from "elophanto/EloPhanto"

Per-company trust state procedure (learning / trial / operating). When a live outreach tool refuses, names the canonical draft replacement.

2026-05-2688

voice-extraction-workflow.md

from "elophanto/EloPhanto"

Produce a voice contract from operator-curated exemplars (Phase 10). Triggered by exemplar paths, voice intent, or draft lint failure.

2026-05-2688

strategy-foundations.md

from "elophanto/EloPhanto"

Marketing-strategy foundations distilled from the Phase 11 strategy generator. Use BEFORE drafting a strategy or proposing tactics so output reflects the PESO / RACE / STDC / JTBD frameworks and avoids generic agency clichés. Pairs with the `company_plan` tool — this skill is the meta-grounding; the tool runs the full LLM strategy generation.

2026-05-2688

x-virality.md

from "elophanto/EloPhanto"

Concrete, algorithm-grounded guidance for writing X posts that the For You feed ranks high. Distilled from the public source of X's ranking system (github.com/xai-org/x-algorithm, May 2026 release). Use this BEFORE drafting any X post or reply. Pairs with the `x_style_preflight` tool (mechanical banned-phrase check) and the `twitter-marketing` skill (general practice).

2026-05-1788

package.json

"author": "elophanto"

"repository": "elophanto/EloPhanto"

View GitHub Repository View Creator Repositories

$ install --global

$ download --local

Run Skill in Manus

$ useful --forSOC

Market Research Analysts and Marketing SpecialistsBusiness and Financial Operations Occupations13-1161L4

name	b2c-marketing-voice
description	Anti-slop principles for B2C social posts and outreach. Use BEFORE drafting an email, X post, or prospect message so the body reads like a person, not an LLM. Pairs with the Phase 10 voice contract (data/companies/<slug>/voice.yaml) — this skill is the meta-principles; voice.yaml is the per-company specifics. Distilled from public marketing patterns (jackfriks/b2c-marketing on ClawHub) — not a copy.
license	MIT
metadata	{"author":"petr-royce","version":"1.0.0","phase":10}

B2C marketing voice — anti-slop rules

Triggers

email_draft
outreach_draft
post_draft
twitter_post
write a tweet
write a post
cold email
outreach message
social media post

The four dead patterns (do NOT use)

Self-focused hook. We help businesses leverage AI to... / Our team is excited to... / At <Company> we believe... — reader scrolls past in 0.3s. Open on THEM or on a CONCRETE OBJECT.
Generic CTA. DM me, check the link in bio, book a call, learn more — these are not CTAs, they are noise. Either soft-CTA (if this resonates...) or zero-CTA (let the post stand).
Corporate filler. leverage, unlock, synergy, in today's fast-paced world, are you tired of, imagine if, let me explain why. These are LLM tells. Any one of them and the reader knows.
Abstract claim with no object. AI is changing everything / productivity is the new luxury. No screenshot, no number, no person — nothing for the reader to anchor on.

The four winning shapes (use one)

Another person + conflict + showed them + changed mind. My dad didn't believe agents could ship real code. I showed him the PR diff. Now he asks me daily what they shipped.
POV: . POV: you check your agent's overnight log and find 4 closed tickets you didn't write.
I used to . Then . Now . I used to think prompts were the moat. Then I watched a 12B model out-plan GPT-4 on tool use with the same prompt. Now I think the orchestration is.
Concrete object up front. A screenshot, a number, a single observed line of code, a single quoted sentence from a user. Then your one-line take. Then stop.

Length defaults (override per voice.yaml)

X post: 80-240 chars. If you wrote 280 you padded. Cut.
Cold email body: 40-80 words. Subject is its own contract — no preamble in body.
LinkedIn post: 400-800 chars. The hook line stands alone; the rest is one block.

Before you draft

voice_show — if has_voice=True, the company's voice.yaml is the final authority and overrides anything in this skill.
Pick one winning shape from the list above. Name the shape in your reasoning so the operator can audit your choice.
Identify the CONCRETE OBJECT or CONCRETE PERSON. If there isn't one, don't draft yet — go find one (call browser_navigate, ask the operator, read a real prospect's bio).

After you draft

voice_lint(text=<body>) — fix any violations before calling the draft tool. The draft tool will lint again; catching it here saves a round-trip.
Reread the opening line out loud. If it sounds like a press release, rewrite it.
Reread the CTA. If it's DM me or check the link, delete it.

What this skill does NOT do

It does NOT replace the per-company voice.yaml. voice.yaml wins on every conflict.
It does NOT apply to B2B technical writing (changelog, docs, internal memos) — those have different conventions.
It does NOT produce hooks for you. It tells you which shapes work and which fail; the operator's exemplars (and your judgement) produce the actual line.

b2c-marketing-voice

B2C marketing voice — anti-slop rules

Triggers

The four dead patterns (do NOT use)

The four winning shapes (use one)

Length defaults (override per voice.yaml)

Before you draft

After you draft

What this skill does NOT do

More from this repository

More from this repository

B2C marketing voice — anti-slop rules

Triggers

The four dead patterns (do NOT use)

The four winning shapes (use one)

Length defaults (override per voice.yaml)

Before you draft

After you draft

What this skill does NOT do