Competitor Tracker

Monitor and analyze competitors' influencer marketing: who they partner with, what campaigns they run, how they structure collaborations, and what results they appear to achieve.

Quick Start

Shortest invocation:

Monitor [competitor name]'s influencer marketing activities

Compare a set of rivals and surface gaps:

Compare influencer strategies across [competitor 1], [competitor 2], and [competitor 3], then show me which influencers they're missing in [category]

Skill Contract

• Reads: your brand name, the competitor set, platforms to monitor, time period, focus areas (partnerships/campaigns/content/all). Public creator handles and post data the user supplies or that ~~social platform analytics returns.
• Writes: a competitive intelligence report saved to memory/influencer/competitor-tracker/YYYY-MM-DD-<topic>.md (partnership roster, campaign analysis, content-strategy review, performance estimates, side-by-side comparison, opportunity list).
• Promotes: durable facts (named competitors, their primary tiers/platforms, confirmed exclusive partners, recurring campaign windows) to memory/hot-cache.md. Competitor-partner and exclusivity flags for creators already on the roster go as one-line updates to memory/events/creators.ndjson via an authorized operation: propose request to registry-events.py for creator-registry to reconcile.
• Done when:

1. Each tracked competitor has a partnership roster and campaign breakdown with sources or stated estimates.
2. A side-by-side comparison table covers your brand plus every competitor.
3. At least 3 ranked opportunities (untapped creators, strategy gaps, or open platforms) are listed.

• Primary next skill: campaign-planner

Handoff Summary

> Emit the standard shape from skill-contract.md §Handoff Summary Format.

Data Sources

This family is Tier 1 — it works with no live integrations. Ask the user for the competitor set, the platforms, and any creator handles they already know, then build the analysis from public posts and stated estimates.

Where a tool could speed things up, use ~~ connector placeholders:

• ~~influencer database — pull a competitor's known partner roster and tier mix.
• ~~social platform analytics — estimate reach, engagement rate, and post cadence per creator.
• ~~CRM — cross-check whether a former competitor partner has already touched your pipeline.

Keyless news read on rivals: python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/gdelt.py" '"<competitor>"' --days 30 lists a rival's global news coverage with no key — campaign launches, partnership announcements, PR pushes — Measured from GDELT's news index (news media only, not social posts; ≥5s between calls). See scripts/connectors/README.md.

Rival-partner channel watch (free key / keyless): python3 "${CLAUDE_PLUGIN_ROOT}/scripts/connectors/youtube.py" channel <partner-handle> reads a competitor partner's real subscriber/view counts (free YOUTUBE_API_KEY), and every YouTube channel also has a keyless RSS feed — https://www.youtube.com/feeds/videos.xml?channel_id=UC… piped into rss_monitor.py — for tracking partner posting cadence and spotting a burst of sponsored content without any API at all.

Label every estimate as an estimate. See CONNECTORS.md for the keyless/free recipe per category.

Instructions

Each step has a fill-in template in references/templates.md.

1. Define competitive set — capture your brand, prioritized competitors (direct/indirect), platforms, time period, and focus areas. (template)
2. Track influencer partnerships — for each competitor, build a current/recent partner roster (handle, platform, followers, partnership type, duration), then the observed selection criteria, relationship-type mix, partnership frequency, and notable partners. (template)
3. Analyze campaigns — break down recent campaigns (timeline, platforms, tier mix, content type, hashtag, CTA, estimated spend, what worked/didn't), plus a calendar and seasonal/launch patterns. (template)
4. Review content strategy — log format preferences, content themes, messaging, hashtag strategy, and creative direction. (template)
5. Estimate performance — estimate overall program metrics, performance by platform and by tier, and top/underperforming content. Mark every figure as an estimate. (template)
6. Generate competitive comparison — a side-by-side table (your brand + every competitor), a strategy-element matrix, and a share-of-voice bar. (template)
7. Identify opportunities — rank untapped and former-competitor creators, strategy gaps, and platform/niche/format openings (at least 3 ranked). (template)
8. Generate insights report — executive summary, strategic recommendations (immediate/short/long-term), tracking recommendations, and next review date. Save to the memory path above. (template)

Worked Example

User: "Track the influencer marketing activities of Glossier, Fenty Beauty, and Rare Beauty"

Output: Competitor analysis showing Glossier's UGC-heavy approach, Fenty's diverse creator network, Rare Beauty's mental health-focused partnerships, with identified gaps and ranked opportunities. Full invocation patterns, "what this skill does", and tips for success live in references/templates.md.

Reference Materials

• references/templates.md — fill-in templates for all 8 steps, invocation patterns, worked example, and tips.
• skill-contract.md — shared contract and handoff summary format.
• state-model.md — memory tiers and save-path conventions.
• CONNECTORS.md — keyless/free data recipe per ~~ connector category.
• Sibling Map skills: influencer-discovery — find creators competitors aren't using; fit-scorer — score competitor partners for your brand.
• trend-spotter — spot trends competitors are riding.

Next Best Skill

• Primary: campaign-planner — turn competitive gaps into a differentiated campaign.
• Alternate (Map): influencer-discovery — pursue the untapped and former-competitor creators this analysis surfaced.
• Alternate (Map): fit-scorer — score a competitor's roster against your brand before you poach.

Termination note: keep a visited-set of skills invoked this session. If the next skill has already run this session, stop and report the chain complete instead of re-invoking. Max chain depth is 3 hops.