What is actually shipped
Every sibling site publishes this page, and the rule is the same everywhere: this is the honest version, not the encouraging one.
The lifecycle table, unsoftened
| Lifecycle stage | For code | For skills |
|---|---|---|
| Version control | everywhere | in git, no versions of their own |
| Dependency management | pip / npm | none |
| Vulnerability management | partially | one review, Feb 2026, never repeated |
| Code review | standard | no process |
| CI | extensive | no skill is tested |
| Distribution | PyPI, Docker Hub | no registry |
| Documentation | standard | the one stage where skills lead |
One green cell out of seven. Full context on the thesis page.
The duplicate
use-sg-playwright exists in two repositories, 43 words apart, and one copy carries an auth claim the service code contradicts. The full exhibit, measured on every build →
Zero declared permissions, zero evals, no registry
- Permissions: 0 of 8 shipped skills declare any. The design exists; nothing implements it.
- Versions: 0 of 8 shipped skills carry a
versionfield. - Evals: zero exist for any shipped skill, despite the scoring position (the economy page) depending on them existing.
- Registry: none. This site’s catalogue is the closest thing to one, and it is a generated list, not a package manager.
The security precedent, and the lapse
In February 2026 — before any of the June theory — the file-transfer skill got a full review trio: an AppSec threat model, a DPO data-flow analysis, and a GRC risk register, all dated 26–27 Feb. That is the “every analogy we have with code, we need with skills” discipline actually practised: a skill treated as software, reviewed as software.
Once. No skill since has had the same treatment. It belongs here as both the precedent that proves the discipline is possible, and the lapse that shows it was not repeated.
Every current skill is a static photograph
The graph projection theory criticises exactly the shape every shipped skill has today: a fixed markdown file rather than a projection of an underlying graph. Publishing that criticism against the estate’s own artefacts is the credible move; resolving it is a research programme, not a sprint.