Google Trends data captured today (July 28) shows rising searches for Claude Code skills that supply has not caught up with. Searches like "task observer skill" and "claude record a skill" are surging this month and match nothing in a catalog that tracks 58,971 skills. Two more risers match skills so small they are effectively invisible. That is demand sitting in public search data, waiting for someone to ship against it.
What the search data says
We pulled the rising queries around "claude skill" and "claude code skills" from Google Trends (worldwide, past month). Three clear buckets came out.
Bucket 1: demand that supply already meets. These rising searches map cleanly to existing, heavily-installed skills:
- "ponytail" (+200%) - ponytail (GitHub), 21,720 installs. The output-compression skill keeps climbing week over week.
- "ui-ux-pro-max-skill" (+140%) - ui-ux-pro-max (GitHub), 287,987 installs.
- "front end design skill claude code" (+120%) - Anthropic's own frontend-design (GitHub), 711,652 installs, the second-most-installed skill anywhere.
- "grill me skill" (+80%) - grill-me (GitHub), 680,248 installs.
Bucket 2: demand with no supply at all. These are risers with zero matching skills in the catalog:
- "task observer skill" (Breakout, Google's label for a query growing faster than 5,000%) - developers want a skill that watches long-running agent tasks and reports on them. Nothing shipped matches it. The only near-name hit is a design-patterns skill about the observer pattern, which is a different thing entirely.
- "claude record a skill" (+350%) - people want to record a workflow and turn it into a skill, the way you record a macro. No skill does this today.
Bucket 3: demand outrunning tiny supply. The searches are Breakout risers, the matching skills exist but have barely been found:
- "i have adhd skill claude" (Breakout) - one matching skill exists: i-have-adhd, an output style that leads every reply with the next concrete action and suppresses tangents. It has 2,799 installs against a Breakout search curve. The demand is visibly bigger than the distribution.
- "dataviz skill" (Breakout) - dataviz-skills exists with 163 installs. A Breakout query against a triple-digit install count is a positioning gap, not a supply gap.
Why this matters if you write skills
The skill ecosystem's biggest install counts went to skills that met an obvious, repeated need: making agents argue back (grill-me), compressing output (ponytail), designing frontends (frontend-design). Every one of those needs showed up in search before it showed up in install counts.
A Breakout search with zero or near-zero matching supply is about the strongest free product signal there is. Whoever ships a competent task-observer skill, or a record-to-skill workflow, or simply out-executes the tiny incumbents in bucket 3, gets the default position for a search that is already growing.
The full rising list refreshes on the Trending in search rail as new Trends data lands.
We run Skillselion, the catalog referenced here. Data: Google Trends (worldwide, past month, captured July 28, 2026) cross-checked against the live catalog, refreshed daily from skills.sh, GitHub and MCP registries, ranked by real installs. Skillselion is independent and not affiliated with Anthropic, OpenAI or Cursor.

Top comments (0)