influencer

Influencer Command

Run the influencer-marketing lifecycle along the STAR loop (Scout → Target → Activate → Report): understand the audience, find and score creators, plan and brief the campaign, run outreach and amplify, then track ROI. Skills score on the STAR framework (Suitability / Trust / Appeal / Return → SQS) and operate from the user's own data and project memory — keyed creator-analytics suites are never required; connectors only automate retrieval.

Route

Infer the phase from the goal (or honor --phase) and route to the matching skill:

• Scout — audience-mapper (audience/niche modes), trend-spotter, influencer-discovery, fit-scorer (STAR Suitability); creator-registry dedupes candidates against the roster
• Target — competitor-tracker, campaign-planner, brief-generator, budget-optimizer
• Activate — outreach-manager, creator-content-auditor (STAR gate), contract-helper, content-amplifier (paid whitelisting / UGC repurpose modes) — resolve the carried opaque creator_ref through an authorized artifact or verified registry link, then consult the creator-registry projection (memory/creators/<aggregate-id>.md: contact path, last agreed rate, exclusivity, compliance history) before outreach or contracting; never derive the path from a raw handle
• Report — landing-optimizer (post-click), performance-analyzer, roi-calculator (STAR Return), report-generator

Rules

• Start where the goal sits in the funnel; do not force the full four-phase chain when the user only needs one stage.
• creator-content-auditor is the pre-publish gate: any creator content goes through its STAR Trust check (FTC disclosure STAR-T1, claim integrity STAR-T2) before it ships.
• For sponsored_content_gate, require disclosure status, claim evidence, and the governing brief. If any applicable evidence is unobserved, keep it Unknown and return NEEDS_INPUT/UNDECIDED/NOT_SCORED; missing evidence is not a veto. One independently verified veto maps to DONE_WITH_CONCERNS/FIX (Revisions Required); two or more map to DONE/BLOCK (Reject/Hold). A business BLOCK never becomes execution status: BLOCKED.
• Return the audit inline by default. Only with explicit exact-write permission, a validator-clean v3 artifact, and a supported runtime writer may class: auditor-output be persisted to memory/audits/influencer/; otherwise identify that intended sink and ask for authorization.
• memory/events/creators.ndjson is the roster history. Other skills submit authorized operation: propose events; creator-registry alone accepts/rejects or mutates canonical creator state. Run it when proposals are pending or a campaign cycle closes; memory/creators/ contains generated views.
• Score creators/content/campaigns on STAR (Suitability/Trust/Appeal/Return → SQS); label every metric Measured / User-provided / Estimated; never fabricate reach or rates.
• Tier 1 by default — works from user-provided data; connectors only automate retrieval. Compliance checks are guidance, not legal advice.
• Follow each skill's Next Best Skill handoff; stop at the documented termination rules rather than auto-chaining the whole discipline.
• Scope edge — creators vs adjacent lanes: "launch a product with creators" starts at campaign-planner while the launch itself runs on RAMP via /aaron-marketing:launch; "boost / repurpose this" is content-amplifier with paid execution handed to /aaron-marketing:ad; the always-on social calendar belongs to ECHO (social-calendar-builder); contract/rate/exclusivity records live in creator-registry, and any email opt-in evidence in consent-registry.
• Paid creator whitelisting order: after the partner is confirmed, run contract-helper first for exact paid usage/whitelisting rights, then brief-generator for the governing brief. Only after the agreement/brief and draft asset exist, run creator-content-auditor (STAR-T1 disclosure + STAR-T2 claim integrity), followed by content-amplifier in its paid-planning mode. Then hand paid execution to /aaron-marketing:ad --phase activate, where ad-account-auditor gates before spend. Stop for explicit spend approval; no contract, creator audit, or amplification plan authorizes activation by itself.