Fetching from the wire…
01
02
03
04
05
06
07
08
Tools2026-08-13 · source-backed
arXiv 2608.10906 collected 3,797,117 SKILL.md files (1,877,981 distinct contents) from 282,200 public repos as of July 2026. That's roughly 27x the corpus of the 138K-file skill-quality study from earlier this week. The authors' framing is right: skills are a new artifact class, natural-language instructions selected probabilistically, with no compiler or type checker to catch defects. This is the substrate for a skill linter or an internal reuse audit, and somebody should build both.
Each link below shares sources, entities, or timing with this story.
Shared entities / Same source / Earlier coverage
Both cover GitSkills, July, SKILL, SQLite; cite the same source (arXiv 2608.10906); earlier GitSkills coverage from 2026-08-12.
Tailscale uses SQLite / Shared entity: July / Earlier coverage
Linked by a graph relationship (Tailscale uses SQLite); both cover July; earlier July coverage from 2026-07-31.
Shared entities / Same source domain / Shared topic / Earlier coverage
Both cover July, SKILL; reported by the same outlet (arxiv.org); overlapping topics (corpu, skill).
OpenClaw uses SQLite / Shared entity: July / Same source domain / Earlier coverage
Linked by a graph relationship (OpenClaw uses SQLite); both cover July; reported by the same outlet (arxiv.org).
OpenClaw uses SQLite / Shared entity: SQLite / Same source domain / Earlier coverage
Linked by a graph relationship (OpenClaw uses SQLite); both cover SQLite; reported by the same outlet (arxiv.org).
Tailscale uses SQLite / Shared entities
Linked by a graph relationship (Tailscale uses SQLite); both cover July, SQLite.
Shared entities / Earlier coverage
Both cover July, Skill, SQLite; earlier July coverage from 2026-07-30.
OpenClaw uses SQLite / Same source domain / Shared topic
Linked by a graph relationship (OpenClaw uses SQLite); reported by the same outlet (arxiv.org); overlapping topics (artifact, author, catch).