Organic
Create organic content scripts in your voice — Reels, TikToks, carousels, static posts.
Shared source: The portable workflow contract lives in
workflows/mb-organic/workflow.md. This Claude skill is the Claude Code shell
over that source.
Shared contract markers: Keep these aligned with the shared source.
Required commands:
mb status --json --peekmb start --jsonmb doctor repair --planmb validate --cross-refs --jsonmb checkpoint --plan --json
Required fact paths:
money_pathmoney_path.objects.proof.qualitymoney_path.objects.channel_strategymoney_path.objects.active_pushvalidation.file_contractscontent_strategycontent_strategy.overall_statecontent_strategy.simple_entry_pointcontent_strategy.layerscontent_strategy.layers.distributioncontent_strategy.layers.channelscontent_strategy.layers.accountscontent_strategy.layers.peoplecontent_strategy.findingsranked_actionsupdatereadinessdrift.itemsrelationship_health.gapscheckpoint.pendingcheckpoint.pending.blockersruntime.codex_cliruntime.claude_code
Approval gates: updates_repairs_migrations, file_writes, checkpoint,
provider_mutation, publishing_or_spend, customer_contact, private_data,
destructive_operations, structured_collection, and
public_issue_or_proposal.
Public/private boundaries: no_secrets, no_raw_provider_exports,
no_raw_transcripts, no_customer_member_data,
no_private_runtime_settings, no_private_dms_or_gated_communities, and
no_raw_finance_legal_records.
Core flow: support plan, video, carousel, static, sales-video-repurpose, or
review.
Marker phrase: plan, video, carousel, static, sales-video-repurpose, or review.
If the operator wants mining or source collection, route to mb-think.
Use content_strategy.overall_state and
money_path.objects.proof.quality before drafting; keep source/privacy
boundaries visible; route coordinated drafts to
pushes/<YYYY-MM-DD-slug>/organic-batch-001.md; Do not publish, schedule,
upload, mutate provider accounts, spend, auto-DM, auto-reply, or contact
customers. Codex support stays read-only planning until runtime smoke proves
organic drafting writes.
Need help? Type /mb-help + your question anytime. If conversation compacts, /mb-help reloads fresh context.
CLI facts first: Run mb status --json --peek from the business repo before
triage, reference-health checks, or content planning. Use its readiness, drift,
and ranked-action facts instead of hand-rolled repo-health probes.
Output destinations and operator vocabulary
When this skill produces a coordinated push (a content sequence with a goal,
audience, and review date), write the wrapping record to
pushes/<YYYY-MM-DD-slug>/push.md (type: push, linked_pushes for inbound
links). One-off scripts and source captures route to documents/transcripts/
or documents/prototypes/ per the artifact-routing rules in the system
architecture doc; they don't need a push wrapper.
If a content push uses a comment-keyword, DM-keyword, reply/link, resource
delivery, or external provider setup plan, write the durable plan to
pushes/<YYYY-MM-DD-slug>/playbooks/<playbook>.md with type: playbook.
The playbook records the plan, approval state, safe provider state, validation
evidence, and outcomes. It does not execute provider mutation.
If core/vocabulary.md defines display words (e.g. terms.push.singular: drop),
speak the operator's word in conversation while still writing engine files.
If the repo still has legacy campaigns/ records, recommend mb doctor and
mb migrate campaigns --plan before creating new push work.
When creating push.md, include the validator-required frontmatter and fill
the values from repo truth or operator answers:
---
type: push
slug: YYYY-MM-DD-slug
kind: nurture
status: planned
health: unknown
goal: { metric: "", target: "", by: YYYY-MM-DD }
owner: ""
audience: ""
offer: core/offers/<offer>/offer.md
promise: ""
---
When creating or first writing a push, ask once:
"Which bet does this push serve, and which existing decision or research supports it?"
Offer active bets and recent decisions and research by name from status facts or files already read.
If this is a measurable trial with a review or cut point and no covering bet,
route to mb-bet before execution.
Accept "none yet" and continue. For a deliberate one-off with no bet, write
linked_bets: [] and a one-line no_bet_reason. On an approved write, put the
chosen repo-relative paths in linked_bets, linked_decisions, and
linked_research on the push; add its path to linked_pushes on each named
bet, decision, and research file. Leave unanswered links empty; do not invent
a reason. Mirror typed links in ## Related links as relative Markdown links.
Use docs/business-connections.md; never infer links from body-only references.
Also link named playbooks and outcomes through linked_playbooks and
linked_outcomes.
Triage
Mining, scraping, competitor research, transcribing or extracting from video ("mine", "scrape", "research competitors", "what are they saying", "transcribe") is research work: say "Mining is research work — /mb-think handles that and saves to your research/ folder. Should I switch you over?" and route there. "Short clips from sales video", "repurpose this VSL" or "social cuts from this sales video" continue with references/sales-video-repurpose.md. "Create", "generate", "write scripts", "make content" continue here. If unclear, ask: "Are you trying to mine competitor content (research) or create new scripts (generate)?"
Pull Latest Updates
For the standard Main Branch update check and failure warning, see references/pull-engine-updates.md. Run it at the start of every invocation.
Then run mb status --json --peek from the business repo and use its
readiness, drift.items, and ranked_actions facts before asking setup or
reference-health questions. Direct file checks below are content-specific, not
repo-health probes.
Mining Lives In /mb-think
This skill assumes mining already happened: users arrive with concepts from research/*-competitor-mine.md, a topic, or their own research. Mining methodology (Visual/Audible/Emotional framework, AI capabilities and limits, why mining flows into reference): references/mining-methodology.md.
If a selected research file has brief_format: grok-8, use its downstream handoff before generating: ICP and audience language for hooks, journey stage for format and CTA, competitive landscape for contrast, brand story for voice, content/assets for proof, and metrics/constraints for the success bar. If the brief names a resource-delivery mechanic, write or update a push playbook as a plan only; do not execute provider mutation.
When organic work is meant to sharpen an offer instead of only create content, load .claude/reference/conversion/offer-sharpening.md. Use the rubric to turn audience language, objections, proof angles, and soft CTAs into durable offer updates through /mb-think; do not treat comments or saves as buyer proof by themselves.
Offer Context Resolution
Before loading reference files, resolve active offer context with
.claude/reference/business-primitives/offer-bet-push-proof.md:
- If a future
mbJSON field exposes active offer state, use it. - Do not treat
.vip/local.yamlas the source of truth for active-offer state. If legacy state exists, confirm the offer with the user instead of silently routing. - If an offer is selected and
core/offers/[offer]/offer.mdexists, load it as the active offer. - If no offer is selected AND
core/offers/exists: ask which offer. - If no
core/offers/folder: usecore/offer.md(single-offer mode) - If the repo does not have
core/, do not infer current offer context from old paths. Runmb doctor repair --plan; usereference/*only as legacy migration input.
Always-core files: soul.md, voice.md, content-strategy.md
Content strategy layers: read core/marketing/distribution-strategy.md,
core/marketing/channels/<channel>.md,
core/marketing/accounts/<platform>-<account>.md, and
core/people/<person>.md when the run names a channel, account, founder, or
weekly content plan that depends on them. Do not copy person voice into every
account file; reference it.
Offer-aware files: offer.md, audience.md
Proof files: company-wide proof in core/proof/testimonials.md,
core/proof/typicality.md, and core/proof/angles/; offer-specific proof in
matching files under core/offers/[active]/proof/. Read older offer
testimonial files as compatibility context only.
Before turning proof into posts, videos, or outcome claims, read
money_path.objects.proof.quality from mb status --json --peek. Treat
generic testimonials, specific outcomes, offer linkage, typicality context, and
outcome feedback as factual signals. Do not describe proof as good,
high-converting, persuasive, or ready to win.
Offer argument: /mb-organic video [offer] "concept" — e.g., /mb-organic video community "morning routine"
If offer specified, it selects the offer for this run only.
First-Time Setup
Requires core/voice.md (always core), plus resolved offer.md and audience.md (offer-aware — checks core/offers/[active]/ first, falls back to core/).
Optional but recommended: core/content-strategy.md — If present,
/mb-organic reads the recognition target, pillars, asset jobs, non-publishing
rules, and strategy links. If the repo has
core/marketing/distribution-strategy.md, channel files, account files, or
core/people/<person>.md, read the relevant ones before drafting. Works
perfectly without them.
Congruence check: If core/operations/funnel/skool-surfaces.md exists, read it. Organic content should echo the same positioning and claims visible on the Skool about page and pricing cards. No contradictions between organic and the landing experience.
CWD-first: If current Main Branch markers exist in CWD, use it. If CWD looks like an old Main Branch repo, run mb status --json --peek and/or mb doctor repair --plan before saved config or discovery. Do not write to old repo structure. Otherwise, run mb status --json --peek and use its repo/readiness facts if available. If status cannot identify a repo, ask the user or run /mb-setup.
Missing files? See references/first-time-setup.md.
Presenting Options (Keep It Tight)
Don't list all modes in chunky blocks. Check for recent mining in research/, recommend ONE path based on context, and mention alternatives in one line. With concepts: "Found recent mining (research/...) with 10 concepts. Recommended: pick a concept and generate a video script. Other modes: carousel "concept", static "concept". Which concept interests you? Or provide your own topic." Without: offer (1) mine first via /mb-think and return, or (2) give a topic and generate directly.
Modes
/mb-organic (Default)
Check for existing mined concepts, pick one, generate scripts.
Check research/ -> Check content-strategy.md and relevant strategy layers (if present) -> Select concept -> Generate -> Output
If content-strategy.md exists and has pillars or asset jobs defined, suggest topics aligned to those pillars when the user has no specific concept in mind.
If no mining exists, prompt: "No mined concepts found. Want to mine competitors first? That's /mb-think — should I switch you over?"
/mb-organic mine (Routes to /mb-think)
Redirect: "Mining is research work now. Routing you to /mb-think for mining — it'll save to research/ and you can come back here to generate scripts from those concepts." Then invoke /mb-think. Before any large Apify mining run, the owner sees an estimated cost and says OK first (apify-setup.md).
/mb-organic video "concept"
Generate Reels/TikTok script from a concept.
/mb-organic carousel "concept"
Generate multi-slide carousel copy from a concept.
/mb-organic static "concept"
Generate single-post caption from a concept.
Output path (all script modes): pushes/YYYY-MM-DD-organic-[offer]-{slug}/organic-batch-001.md (include offer slug in multi-offer mode; omit [offer]- in single-offer mode). On legacy repos that still have campaigns/, run mb migrate campaigns --plan first; do not write new content under campaigns/.
Output frontmatter:
---
type: output
format: video | carousel | static
date: YYYY-MM-DD
status: draft
platform: instagram | tiktok
---
Campaign name is REQUIRED. Ask user if not provided. Examples: january-hooks, transformation-series, pain-point-reels.
Context Awareness (Check Before Recommending)
At session start, scan what's been done:
- Check
research/*-competitor-mine.md— Who was mined? When? - Check
pushes/*-organic-*/— What scripts exist? Treatcampaigns/as a legacy fallback only. - Don't suggest re-mining same handles from today
- Recommend generating from existing mining if concepts unused
- Check
core/content-strategy.md— What recognition target, pillars, asset jobs, and strategy links are defined? - If the channel/account/person is named, check the matching
core/marketing/andcore/people/files before drafting.
Offer remaining concepts, mining new competitors via /mb-think, or a fresh topic, and ask what to call the batch (examples: january-hooks, transformation-reels). More examples: references/examples.md.
Transparency
Before generating: show which reference files you're using. Before saving: show file paths.
Full Flow Walkthrough
- Check for concepts in
research/*-competitor-mine.md; if none, ask whether to mine first (route to/mb-think) or take a topic directly. - User picks a concept or provides their own; map it to their offer, audience and voice.
- Generate with the matching framework (video, carousel, static).
- Save to
pushes/YYYY-MM-DD-organic-{slug}/, nevercampaigns/(legacy compatibility only). - Checkpoint prompt: run
mb checkpoint --plan --json, show the proposed checkpoint and blockers, validate the chosen message withmb checkpoint --validate "..." --json, then after operator approval save withmb checkpoint --message "..." --yes.
Video Mode
Input: concept from mining, user topic, or research file.
| Framework | Structure | When to Use |
|---|---|---|
| Educational | Hook -> Tips -> Takeaway | How-to, lists |
| Story-based | Hook -> Trigger -> Outcome | Personal narrative |
| Transformation | Before -> Turning Point -> After | Journey, case study |
| Problem-Solution | Hook -> Problem -> Solution | PAS for organic |
Structure: Hook (0-3s) → Retain (3-45s) → Reward (final 5-15s)
See references/organic-frameworks.md and references/video-script-template.md.
Carousel Mode
7-10 slides: Hook → Value (one idea/slide) → Summary → CTA
See references/carousel-template.md.
Static Mode
Hook (first line) → Body → Soft CTA → Hashtags (optional)
See references/static-template.md.
Voice Adaptation
Read current voice context. If CWD looks like an old Main Branch repo, run mb status --json --peek and/or mb doctor repair --plan before relying on it. Match tone, use their vocabulary, avoid their "never say" list.
Authenticity: Sounds like creator (not copywriter). Uses contractions. Matches energy. No AI tells ("dive into", "unlock", "game-changer").
See references/organic-frameworks.md for soft CTA examples.
Integration with /mb-think
To save winning angles: route to /mb-think codify → core/proof/angles/.
Content Strategy Integration
If core/content-strategy.md exists, /mb-organic uses it to improve output:
- Pillar alignment: Suggest topics from defined pillars when user has no specific concept
- Asset job: Identify whether the piece should rank, teach, prove, compare, convert, announce, or start a conversation
- Platform format: Default to the format matching the target platform from channel or distribution strategy
- Account fit: Use account strategy for allowed topics, CTA path, cadence, and content mix when present
- Person voice: Use
core/people/<person>.mdfor founder/person beliefs, stories, and fabrication boundaries when an account references it - Content mix: Track which pillar types have been generated recently to maintain ratio balance
- Hooks library: Pull proven hooks from the hooks library section if populated
- Enemy-pillar mapping: Each pillar fights a named enemy. Content should position against the enemy. Check voice.md for declared enemies.
- Saves-first metrics: Weight save-ability highest when evaluating content concepts. Saveable = educational, actionable, reference-worthy.
If content strategy files do not exist, /mb-organic works exactly as before -- from mined concepts or user-provided topics. No warnings, no degradation.
Quality Checklist
Content: Hook stops scroll. One idea. Value before ask. Soft CTA.
Saves optimization: Is this content saveable? Educational, actionable, reference-worthy content drives saves. Saves are the #1 purchase intent signal — weight them above shares and likes.
Enemy clarity: Does this content fight a named enemy? Check voice.md for Named Enemies section. Enemy-driven content creates identity contrast.
Voice: Sounds like creator. Matches energy. Uses vocabulary. No AI tells.
Platform: Appropriate length. Correct structure. Optimized for retention/saves.
Skool congruence: If skool-surfaces.md exists, claims and positioning match live about page + pricing cards.
Examples & Compaction Recovery
For walkthrough examples (context-aware start, video-only, mining redirect) and the recovering-from-compaction protocol, see references/examples.md.
References
Setup: first-time-setup.md, minimal-voice-template.md, apify-setup.md
Frameworks: organic-frameworks.md
Sales video repurposing: sales-video-repurpose.md
Templates: video-script-template.md, carousel-template.md, static-template.md
Methodology & Examples: mining-methodology.md, examples.md