← ClaudeAtlas

angle-maplisted

Map the reasons-to-buy (angles) and buyers (personas) every competitor is running on Meta, and find the angle nobody owns. Use for "what angles are competitors using", "positioning gaps", "white space", "how do we stand out in ads", "who are they targeting".
seercraft-io/seer-skills · ★ 0 · AI & Automation · score 70
Install: claude install-skill seercraft-io/seer-skills
# Angle map Done means: `seer/reports/angles-<date>.md` with a grid of angle × brand, the crowded angles, the empty ones, and the one empty angle the person's brand could credibly own. Read `${CLAUDE_SKILL_DIR}/../seer-start/reference/house-rules.md` first. `$S` = `${CLAUDE_SKILL_DIR}/../seer-start/scripts`. ## Steps 1. `python3 $S/ideas.py seer/ads/<rivals>-<date>.json --json seer/digest.json` 2. For every live idea with copy, write one angle (the main reason to buy it gives) and one persona (who it speaks to), each in 2-4 words. Then merge near-duplicates into 6-12 angles that fit this category (e.g. for wallets: "slim / less bulk", "lasts forever", "anti-theft / RFID", "the gift", "tech that does more", "status / design"). Keep a list of which ideas got which tag. 3. Build the grid: rows = angles, columns = brands, cells = live ideas (median days running). 4. Crowded = 3+ brands lean on it. Empty = no rival has more than one idea on it. 5. Read `seer/brand.md`. For each empty angle, ask: can this brand truthfully claim it, and do its buyers care? Pick one, and say what proof the brand would need to show. ## Output - The grid (Markdown table), with "tagged N of M ideas" under it. - **Crowded**: angles everyone runs, with the longest-running example of each (cited). - **Owned**: angles one rival clearly owns (most ideas, longest running), cited. - **Empty**: angles nobody runs, marked as untested, not as "better". - **Your angle**: the one to try, the proof