{"version":"2.1.287","anchor":"skill-mentions-attachment-for-human-typed-prompts","canonical_anchor":"skill-mentions-attachment-for-human-typed-prompts","heading":"Claude Code now notes which skills you name in your prompt","tier":"notice","area":"Skills","scope":null,"heads_up":null,"url":"https:\/\/changelogs.core-directive.com\/v\/2.1.287\/e\/skill-mentions-attachment-for-human-typed-prompts","release_url":"https:\/\/changelogs.core-directive.com\/v\/2.1.287","markdown":"### Claude Code now notes which skills you name in your prompt\n\nWhen you name a skill in a prompt you typed, Claude Code now attaches a note listing the skills you named\n\n**Unclear.** It is not clear what the further check tests, or whether the note is meant to push the model toward using the named skills.\n\n**What**\n\nA skill is a packaged set of instructions Claude can load for a particular kind of task, and the Skill tool is how it loads one. When you type a prompt yourself, Claude Code now checks it for skill names and attaches a note listing the skills you mentioned. This happens only when:\n\n- the prompt was typed by a person, not generated\n\n- the Skill tool is available in the session\n\n- a further internal check passes\n\nThe check reads your prompt as you typed it, before any of it is expanded.\n\n**Why**\n\nIn practice, the model now gets a separate note saying which skills you named, alongside your prompt.\n\n- Area: Skills\n- Tier: You'll notice\n- Useful: 2\/5\n- Signal: 2\/5"}