Plan de validation

Ce que l’Agent doit faire ensuite

1. 1Inspect repository, README/SKILL.md, license, and recent commits before production use.
2. 2Install in an isolated workspace or sandbox with no production secrets available.
3. 3Run the smallest representative task and record files touched, commands run, network access, and outputs.
4. 4Compare the selected skill against at least one alternative when the eval status is review or failed.
5. 5Promote only after the agent reports a successful verification result and unresolved warnings are accepted.

Ne pas utiliser quand

Conditions exigeant une autre skill

• Équipes qui nécessitent un SLA soutenu par le fournisseur
• production agents without a repository review
• Low GitHub adoption signal
• No executable scripts are included in the skill itself; it depends on external fistPkg and other skills, which limits standalone verification but is not a security risk.
• The SKILL.md excerpt appears truncated in the provided documentation, though the repository likely contains the full file; ensure the complete workflow and limitations are present.
• References to JTG standards are provided as checklists and explicitly avoid reproducing copyrighted text, which is good, but users must still confirm they have legal copies of the standards.