wiki

The skill documentation and workflow are coherent with the stated purpose (extracting domain knowledge, creating a git-backed wiki, and packaging it as an agent skill). However, multiple supply-chain and autonomy risks are present: (1) it recommends creating a .claude settings file to bypass permission prompts, enabling more autonomous actions; (2) it encourages transitive installation via `npx skills add` without guidance on vetting or pinning; and (3) the extraction + automatic commit/push flow can inadvertently exfiltrate sensitive files if remotes are misconfigured or secrets are not filtered. These patterns make the skill medium-risk from a supply-chain and data-exposure perspective. Mitigations: do not enable bypassPermissions by default, implement secret detection/exclusion in extraction, require explicit user confirmation before any git push to remote URLs, and recommend pinned, vetted skill packages with signature checks for distribution.