Content Amplifier

Content Amplifier

Extract more value from a frozen, approved creator asset. Two modes: paid (extend reach with paid spend — whitelisting, Spark Ads, dark posts, budget + optimization) and repurpose (reuse one asset across paid, website, email, and social — inventory, repurposing map, format specs, distribution plan). Both start from the exact version cleared by creator-content-auditor and active scoped rights. Spark/boost paths additionally require the platform state and live post they depend on; a dark post need not have been published organically.

Scope guard: this skill does NOT score a deliverable for brand alignment, message accuracy, or FTC/disclosure compliance, and it does NOT compute a STAR Trust/Appeal score or run the STAR-T1/STAR-T2 veto — that is the creator-content-auditor gate's job. This skill works the downstream lever: turning approved content into paid reach or many-channel assets, then hands off. In a product launch, this skill owns the repurposing map and the paid-amplification / distribution execution calendar (including the 30-day plan for launch content); the launch discipline's momentum-planner schedules only the launch moments and hands the distribution work here. In always-on organic social the split is the same shape: the standing brand posting calendar belongs to social-calendar-builder and net-new idea-to-multi-platform package drafting to social-creative-builder — this skill keeps repurposing of existing assets and ALL paid amplification, and the social discipline only flags boost-worthy organic winners to it.

Mode selector

Pick with --mode paid or --mode repurpose. If unset: "boost / amplify / whitelisting / Spark Ads / dark post / paid spend / budget" → paid; "repurpose / reuse / turn one video into many / asset library / social proof on pages / multi-channel rollout" → repurpose. If the request spans both (e.g. "cut ad variations and plan the paid spend"), run repurpose first to produce the assets, then hand to paid — do not silently merge; state which mode you ran.

Quick Start

Shortest invocation:

Which frozen auditor-approved asset with current active paid rights should we amplify from [campaign]?  # paid
How can we repurpose this frozen approved asset within its evidenced active destination scope?           # repurpose

Common scenarios:

--mode paid: Create a $5,000 TikTok/Instagram plan from these frozen approved assets and supplied current active scoped-rights records
--mode repurpose: Build a 30-day plan from these 3 frozen approved videos and supplied current active rights records covering the requested destinations

Output expectation — paid: every evidence-complete candidate scored, tiered, and given a spend that sums to budget, plus a scale/pause playbook. Any candidate missing organic-performance evidence or a Hook/Message/Quality/CTA observation remains NOT_SCORED/NEEDS_INPUT with no /25, rank, tier, or spend. repurpose: every source asset rights-tagged, at least one eligible asset mapped to 3+ formats across 2+ channels, plus a dated distribution plan.

Skill Contract

• Reads:
  • paid — stable opaque creator_ref values plus transient creator locators, platform, content type, dated organic reach/engagement/views evidence, separate dated Hook/Message/Quality/CTA observations, amplification budget, campaign objective (awareness/traffic/conversions), target platforms, any prior performance data the user provides, and each candidate asset's current rights status/evidence plus existing placements when present.
  • repurpose — source UGC assets (videos, reels, reviews, images), stable opaque creator_ref values plus transient creator locators and platforms, exact frozen approved_asset_ref, opaque authorized source_ref, usage rights per asset, rights status with observation time/evidence, original performance metrics, target channels, and existing placements. For atomizing a source, the pasted transcript/caption/review text is transient evidence bound to that three-ref source identity; a raw handle or content URL never becomes source_ref.
  • Both pull prior campaign context from memory/hot-cache.md when memory-management is active.
• Writes: return the mode's deliverable and handoff inline by default; save to memory/influencer/content-amplifier/YYYY-MM-DD-<topic>.md only with exact WARM-save authorization. Saved artifacts and handoffs reuse an explicitly carried opaque creator_ref, or a verified creator-registry aggregate ID; otherwise generate one random creator-<UUIDv4> once for the lineage. Never persist a raw handle, creator name, profile/content URL, email, provider ID, or deterministic hash as identity. Keep only creator_ref, frozen asset/approval refs, opaque evidence/rights/contact/placement refs, and non-identifying campaign data; every saved atom source must carry creator_ref + exact approved_asset_ref + opaque source_ref. If no authorized resolver exists, mark identity unresolved and require the transient locator again when needed.
• Promotes: only with separate exact authorization, promote durable facts — paid: chosen amplification mix, per-creator spend tiers, winning audiences, scale/pause thresholds; repurpose: rights levels, expiration dates, library naming convention, top-performing source assets — to memory/hot-cache.md. A newly observed revocation, dispute, expiry, or other rights-status change is not made canonical here: offer only a separately authorized handoff to the existing creator-registry proposal path with the exact status evidence. This skill does not append or accept that proposal.
• Done when:
  • paid — (1) each evidence-complete candidate has dated evidence for organic performance plus Hook/Message/Quality/CTA and is scored /25 before any tier or spend; every incomplete candidate is NOT_SCORED/NEEDS_INPUT with no total/rank/tier/spend; (2) a budget allocation across only evidence-complete eligible content, objectives, and platforms sums to the stated budget; (3) an optimization plan with supplied KPI targets and scale/pause rules is recorded, or those rules stay NEEDS_INPUT; (4) every selected asset has current active rights evidence covering the intended use.
  • repurpose — (1) every source asset has a rights level, expiration, active | expired | revoked | disputed | unknown status, status_observed_at, and status_evidence_ref; (2) at least one source asset is mapped to 3+ distinct output formats across 2+ channels; (3) a dated distribution plan with an asset checklist exists; (4) any existing placement affected by a non-active or out-of-scope status appears in the manual removal queue.
• Primary next skill: paid → performance-analyzer once campaigns are live; repurpose → landing-optimizer to place the repurposed social proof where it converts.

Handoff Summary

Emit the standard shape from skill-contract.md §Handoff Summary Format. State which mode ran. Label every metric Measured / User-provided / Calculated / Estimated / Unknown. Never fill a missing CPM, ROAS, view count, rights date, or decision threshold with an example value: ask for the export or use Unknown/NEEDS_INPUT. Organic performance and the Hook/Message/Quality/CTA observations are selection evidence, not forecast fields: if any is missing, do not substitute Estimated; set that asset to NOT_SCORED/NEEDS_INPUT and emit no /25, rank, tier, or spend. Use Estimated elsewhere only when the user supplied or approved the calculation inputs and assumptions, and show that basis.

Data Sources

This family is Tier 1: both modes work with no live integrations. Ask the user for the mode's inputs and produce the supported portion of the artifact from those. Never invent reach, engagement, CPM, ROAS, rights numbers, or selection observations. Missing organic reach/engagement/views evidence or Hook/Message/Quality/CTA evidence is never Estimated; it makes that asset NOT_SCORED/NEEDS_INPUT. Other forecast fields may be Estimated only from supplied or explicitly approved inputs and assumptions; otherwise use Unknown/NEEDS_INPUT.

Where a connector could sharpen the output (all optional, opt-in Tier 2/3):

• ~~social platform analytics — pull organic reach, engagement rate, and view counts (both modes) instead of asking the user to paste them.
• ~~ad platform (Meta Ads Manager, TikTok Ads Manager, Google Ads) — read live CPM/CTR/CPC/ROAS for the paid optimization playbook, and confirm Spark Ads / Partnership Ad authorization status.
• ~~influencer database — verify creator audience demographics for lookalike targeting (paid); transiently resolve creator locators while retaining only creator_ref, platform, and contract-rights refs in saved repurpose artifacts.
• ~~DAM / asset library — store and tag processed assets; enforce the naming convention (repurpose).
• ~~CRM — supply retargeting/exclusion audiences (paid); reconcile creator records with usage-rights expirations (repurpose).

See CONNECTORS.md for the verified free/keyless recipe per category. None are required; absent a connector, the user supplies the numbers.

Instructions

Select the mode first (see Mode selector), then run that mode's steps. Each step has a fill-in template in references/templates.md — produce the populated artifact, do not skip the table.

Mode: paid

1. Assess available content — build the content inventory: campaign, piece count, budget, creator_ref, platform/type, and dated organic reach/ER/views values with evidence refs. Do not create an organic-performance score from an absent, undated, or unevidenced metric. Paid Step 1 template.
2. Select content for amplification — record separate 1–5 observations for organic performance, Hook, Message, Quality, and CTA, each with source_ref and observed_at. Only when all five are present may the asset receive /25, rank, tier, or spend. If any component or its evidence is missing, preserve supplied observations, set score_state: NOT_SCORED and execution NEEDS_INPUT, list the exact gaps, and do not estimate, normalize, prorate, or hand-calculate a partial total. Paid Step 2 template.
3. Develop amplification strategy — describe three operational methods neutrally: whitelisting / Partnership / Spark Ads use the creator identity and require matching platform authorization; brand-account ads use the brand identity and require the licensed asset in the brand ad account; dark posts are unpublished ad-account placements. Do not claim that one method improves engagement, preserves authenticity, boosts credibility, or performs better/worse without supplied comparative evidence. Treat method choice and budget mix as a declared test or user decision. Paid Step 3 template.
4. Set up targeting — primary lookalike off the creator's engaged audience, plus expansion segments (interest/behavioral/demographic for awareness; retargeting/custom/lookalike for conversions), ad sets per platform, and exclusions. Paid Step 4 template.
5. Allocate budget — split the stated budget only across rights-eligible, fully scored assets, then by objective and platform; set a pacing schedule (learning → optimization → scaling). Allocations must sum to the stated budget. If no fully scored eligible asset remains, return NEEDS_INPUT and no hypothetical allocation. Paid Step 5 template.
6. Optimization playbook — KPI table (CPM, CTR, CPC, CVR, ROAS) with below/above-target actions, an optimization schedule, A/B tests, and explicit scale-up / pause / creative-refresh thresholds. Paid Step 6 template.
7. Platform-specific setup — verify the frozen asset version is auditor-approved and its rights tracker says active with a dated evidence reference; separately confirm the grant is unexpired and scoped to the intended platform, territory, format, and paid use. expired, revoked, disputed, or unknown fails closed for new use. Spark/boost or other existing-post methods also require the matching live post and platform authorization; dark-post setup does not require prior organic publication. Missing approval, required live state, or usable rights stops activation with NEEDS_INPUT; none may be inferred. Paid Step 7 guide.

Return the populated artifact inline. Offer its exact WARM path for save authorization, then ask separately before any HOT promotion. Any ad-account setup, upload, launch, or spend is a further external mutation requiring exact action approval.