instagram-creator-discovery

name	instagram-creator-discovery
description	Discover Instagram creators through search, hashtags, tagged mentions, posts, and Reels, then enrich and rank them into research pools and shortlists.
metadata	{"postplus":{"familyId":"instagram","familyName":"Instagram"}}

Instagram Creator Discovery

Follow shared public skill rules in:

• postplus-shared public skill rules

Use this skill when the user wants to:

• find Instagram creators or influencers in a niche
• search under follower-band constraints such as 5k-20k
• find creators based on what they post, not only on bio keywords
• build creator pools for outreach or partnership research
• compare creators surfaced from search, hashtag, tagged, or content evidence

Read these references before implementation:

• postplus-shared research preferences
• skills/10-routing/creator-discovery-router/references/candidate-schema.md
• skills/10-routing/creator-discovery-router/references/instagram-candidate-mapping.md
• skills/20-research/instagram-references/actor-selection.md
• skills/20-research/instagram-references/normalized-schema.md
• skills/20-research/instagram-references/tool-contracts.md

Core Rule

Do not default Instagram creator search to profile lookup only.

If the request cares about:

• follower bands
• recent activity
• topical fit
• audience fit

prefer content-first or mixed discovery:

1. collect matching posts / Reels / hashtag results / tagged posts
2. extract candidate usernames from the matched content
3. enrich candidate profiles
4. optionally enrich recent posts for stronger evidence
5. rank creators by profile + content evidence
6. return a research pool plus shortlist

Primary Hosted Collection Keys

• instagram-search
• instagram-hashtags
• instagram-profiles
• instagram-posts

Released enrichment:

• hosted collection key instagram-email-search

Local Tools

• skills/20-research/instagram-tools/scripts/build_instagram_actor_input.mjs
• skills/20-research/instagram-tools/scripts/run_instagram_actor.mjs
• skills/20-research/instagram-tools/scripts/normalize_instagram_dataset.mjs
• skills/20-research/instagram-tools/scripts/extract_instagram_candidate_usernames.mjs
• skills/20-research/instagram-tools/scripts/rank_instagram_creators.mjs

Route Guidance

Use handle-first when:

• the user already has handles
• the user has competitor account lists
• the user wants profile enrichment more than discovery

Use content-first when:

• the user wants active niche creators
• the user cares about follower bands
• the user cares about audience fit from actual content

Use mixed when:

• you need search or hashtag recall first
• then need profile enrichment and creator scoring

V1 Workflow

1. Build actor input from a brief

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/build_instagram_actor_input.mjs \
  --brief <work-folder>/.postplus/instagram-brief.json \
  --actor instagram-search \
  --output <work-folder>/.postplus/instagram-search-input.json

2. Run the collection actor

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/run_instagram_actor.mjs \
  --collection-key instagram-search \
  --input <work-folder>/.postplus/instagram-search-input.json \
  --output <work-folder>/.postplus/instagram-search-raw.json

3. Normalize the dataset

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/normalize_instagram_dataset.mjs \
  --input <work-folder>/.postplus/instagram-search-raw.json \
  --actor instagram-search \
  --dataset-type posts \
  --output <work-folder>/.postplus/instagram-search-normalized.json

4. Extract candidate usernames

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/extract_instagram_candidate_usernames.mjs \
  --input <work-folder>/.postplus/instagram-search-normalized.json \
  --route content-first \
  --output <work-folder>/.postplus/instagram-candidate-usernames.json

5. Enrich profiles

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/run_instagram_actor.mjs \
  --collection-key instagram-profiles \
  --input <work-folder>/.postplus/instagram-profile-input.json \
  --output <work-folder>/.postplus/instagram-profiles-raw.json

Then normalize:

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/normalize_instagram_dataset.mjs \
  --input <work-folder>/.postplus/instagram-profiles-raw.json \
  --actor instagram-profiles \
  --dataset-type profiles \
  --output <work-folder>/.postplus/instagram-profiles-normalized.json

6. Rank creators

node ${CLAUDE_SKILL_DIR}/../instagram-tools/scripts/rank_instagram_creators.mjs \
  --profiles <work-folder>/.postplus/instagram-profiles-normalized.json \
  --content <work-folder>/.postplus/instagram-search-normalized.json \
  --candidates <work-folder>/.postplus/instagram-candidate-usernames.json \
  --route content-first \
  --output <work-folder>/.postplus/instagram-creator-ranking.json

Good Output

Return:

• research_pool
• shortlist
• why each top creator was surfaced
• creator type estimate
• topical fit
• audience fit
• follower band fit
• public contact signals if available

Handoff

Escalate to instagram-account-research when:

• the user already has a stable shortlist and wants deeper account snapshots

Escalate to instagram-content-benchmark when:

• the user wants deeper analysis of the strongest posts or Reels

Escalate to instagram-audience-voice when:

• the user wants comment-language extraction from the shortlisted content