acquire-project-skillslisted

Make a target project's own skills usable in the current session. Reads <TARGET_DIR>/.claude/skills/, symlinks each skill into ~/.claude/skills/ (so Claude Code hot-reloads them on the next turn), and records them in acquired-project-skills.json so the SessionEnd hook can remove them on exit. Invoke this at the START of working in any TARGET_DIR that ships its own .claude/skills/ — including from the project-builder, develop, revise-brief, and define-use-case entry points — or run it standalone to pull in one project's skills. Skips (never clobbers) any name already present in ~/.claude/skills/ or any name that matches a workspace skill.
HaroldHormaechea/project-builder · ★ 5 · Web & Frontend · score 77

Install: claude install-skill HaroldHormaechea/project-builder

# Acquire Project Skills Target projects can carry their own skills under `<TARGET_DIR>/.claude/skills/`. Because this Claude session is launched from the **project-builder workspace** (not the target), those project-local skills are not loaded automatically. This skill bridges that gap: it symlinks them into the user's personal skills folder (`~/.claude/skills/`), which Claude Code watches and hot-reloads, so the target's skills become callable in this session without a restart. Every symlink is logged in `<SESSION_DIR>/acquired-project-skills.json` (gitignored), keyed by `(project, session_id)`. The `SessionEnd` hook removes exactly this session's links on exit; nothing leaks into the next session. ## When to run - At the **start of work** on a TARGET_DIR — the `project-builder`, `develop`, `revise-brief`, and `define-use-case` entry points should invoke this once `TARGET_DIR` is resolved, before doing their real work, so any project-shipped skills are available. - **Standalone**, when the user asks to "use / load / pull in the skills from project X." - Re-running it is safe and idempotent: an already-correct link is left in place; nothing is duplicated. A freshly scaffolded project usually has no `.claude/skills/` yet — in that case this is a no-op and you can move on. ## Step 1 — Resolve folders 1. Run `pwd` via Bash to confirm `SESSION_DIR` (the workspace; this is where the script and ledger live). 2. Determine `TARGET_DIR` (absolute path): - If an invoking ent