{"version":"2.1.315","anchor":"project-scope-imports-use-a-sandboxed-filesystem-path","canonical_anchor":"project-scope-imports-use-a-sandboxed-filesystem-path","heading":"Project imports can use a safer way of writing files","tier":"internal","area":"Elsewhere","scope":"individual","heads_up":false,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.315\/e\/project-scope-imports-use-a-sandboxed-filesystem-path","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.315","markdown":"### Project imports can use a safer way of writing files\n\nImporting memory files, skills, commands and settings into a project can now use safer file writing, and reports links and permissions not copied\n\n**Unclear.** It is not stated what condition switches the safer writing on, so whether it applies to a given import is unknown.\n\n**What**\n\nWhen Claude Code imports memory files, skills, commands or settings into a project, meaning the files that belong to the current project rather than your whole account, it can now use a safer way of writing them instead of plain writes. This only happens when a particular condition inside Claude Code is met.\n\nWhen copying skills, Claude Code now reports:\n\n- symbolic links (shortcut files that point somewhere else) that it did not copy\n\n- files whose setuid, setgid or sticky bits, which are special file permissions, were not kept\n\nIf a copy stops partway, it says that a partial copy may remain and where.\n\n**Why**\n\nThis is aimed at stopping an import from writing outside the project through a symbolic link. You are also told what was left out of a copy.\n\n- Area: Elsewhere\n- Tier: Under the hood\n- Useful: 1\/5\n- Signal: 2\/5\n- Scope: individual\n- Heads-up: no"}