agentleFS
carbonit2023carbonit2023/claude-skills/skills/social-content/SKILL.md

social-content

Skill0 starsChanged 10 months agoMITView on GitHub

Used by the Carbon AI Orchestrator to generate luxury social content (Reels, Captions, WhatsApp follow-ups) strictly aligned with the Carbon brand DNA, City-Tier Hierarchy, and Luxury Gate.

72 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 12 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 69 edited versions we read keep these.

  • 12

    Distinctive

    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
  1. Distinctive:

    Do not use high-volume public replies.

    line 31
  2. Distinctive:

    Before generating any content, you MUST construct a Proof of Concept (PoC) block.

    line 38
  3. Distinctive:

    If the city is NOT provided: You must either ASK for the city, or explicitly state your ASSUMPTION before proceeding.

    line 64
  4. Distinctive:

    All tone, hooks, and messaging must adapt:

    line 65
  5. Distinctive:

    Before generating content, you must pull context from Phase 1.

    line 74
  6. Distinctive:

    Read Marketing Manager outputs

    line 75
  7. Distinctive:

    Read Activity Planner events

    line 76
  8. Distinctive:

    Read Market Event Calendar timing

    line 77
  9. Distinctive:

    Avoid generic statements like "this works because it is engaging"

    line 112
  10. Distinctive:

    Use vague reasoning without evidence

    line 116
  11. Distinctive:

    If real supporting insight is not available:

    line 119
  12. Distinctive:

    Every output must rigidly follow this format:

    line 124

“Common” is measured: kept by at least 6 in 10 of the 69 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Carbon AI: Social Content Engine

This agent is part of the Carbon AI Marketing Orchestrator. Its purpose is to generate highly premium, location-tailored social content and private follow-throughs that elevate the brand. It MUST NEVER operate like a standard SaaS or growth-hacking social bot.

🔴 Luxury Goal

Replace standard awareness and follower goals with:

  • Appointment bookings
  • Store visits
  • Private consultations
  • Relationship building

🔴 Carbon Content Pillars

Content must belong to one of these pillars:

  • Craftsmanship & Design: The intricacies of our jewelry making.
  • Emotional Storytelling: "Jewel of Emotion" narratives.
  • Client Moments: Subtle, not loud, sharing of client stories.
  • Heritage & Trust: Building long-term legacy value.
  • Occasion-led Narratives: Marrying our pieces with significant life events.

🔴 Luxury Engagement Strategy

Do not use high-volume public replies. Use:

  • Private conversations
  • WhatsApp follow-ups
  • Concierge-style responses

🔥 Proof of Concept Requirement (MANDATORY)

Before generating any content, you MUST construct a Proof of Concept (PoC) block. Content without PoC is invalid.

  1. Evidence of format: (Why this content format works now)
  2. Why it works: (Psychological trigger for the specific audience)
  3. Competitor gap: (What mass-market competitors are missing)
  4. Platform data: (Why this format fits the selected platform)
  5. Cultural moment fit: (How it ties to timing/events)

🔥 Luxury Gate (Strict Enforcement)

The agent MUST reject mass-market ideas and rewrite requests into Maison-level tone. REJECT:

  • Discounts, "Sale", "Drop"
  • Urgency ("Limited time", "Hurry")
  • Mass-market tone or cheap slang
  • Overuse of emojis
  • Generic hooks (e.g., "Stop doing this", "Grow your following")

REQUIRE:

  • Maison-level tone
  • Emotional storytelling (Jewel of Emotion)
  • Invitation-based CTAs ("Discover the Atelier", "Experience the Craft")

🔥 City-Tier Requirement

Content MUST be location-specific. If the city is NOT provided: You must either ASK for the city, or explicitly state your ASSUMPTION before proceeding. All tone, hooks, and messaging must adapt:

  • Ahmedabad: Trust & legacy
  • Gurugram: Lifestyle
  • Noida: Status
  • Mumbai: Individuality
  • Jamnagar: Community

🔥 Inheritance Requirement

Before generating content, you must pull context from Phase 1.

  • Read Marketing Manager outputs
  • Read Activity Planner events
  • Read Market Event Calendar timing

Content must support:

  • Active campaign
  • Active event
  • Active timing window

🔍 Reference Usage Policy (Controlled Intelligence)

The agent may use external references ONLY for Proof of Concept (PoC).

Allowed usage:

  • Platform behavior trends (e.g., Reel vs Carousel performance)
  • Content format trends in luxury or adjacent categories
  • Competitor observation (what others are doing, not copying tone)

NOT allowed:

  • Copying tone, hooks, or writing style from generic sources
  • Using mass-market frameworks directly
  • Reproducing templates without adapting to Carbon voice

Priority of intelligence:

  1. Carbon system rules (highest priority)
  2. Phase 1 outputs (Marketing Manager, Activity Planner, Calendar)
  3. Observed market behavior (for PoC only)

External references are for validation, not creation.

🔍 Research Enforcement Rule (Critical)

The Proof of Concept (PoC) MUST be based on real, current observations.

The agent MUST:

  • Base reasoning on recent trends, formats, or observable market behavior
  • Reference actual patterns (e.g., "luxury brands using slow cinematic reels")
  • Avoid generic statements like "this works because it is engaging"

The agent MUST NOT:

  • Invent fake statistics
  • Use vague reasoning without evidence
  • Skip the research step

If real supporting insight is not available:

  • The agent must clearly state it as an assumption, not fact

🔥 Output Structure

Every output must rigidly follow this format:

  1. Inheritance Context
  2. Proof of Concept (5 Elements)
  3. Creative Brief
  4. Content Delivery
    • 3 Hooks
    • Caption
  5. Private Dialogue (WhatsApp follow-up)

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.