Menú
Use when adding Claude Code structure to an existing project. Auto-detects stack, creates only missing files.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
Use when adding a feature to a production project. Full cycle: brainstorm, plan, implement, docs.
Use when fixing a bug in a production project. Full cycle: diagnose, fix, regression test, review, docs.
Use to batch-validate all skills in this plugin. Runs deterministic repo-wide static checks (frontmatter, command delegation, cross-references, Architect model= rule, orphan references, README<->MemPalace sync) and reports pass/fail per skill.
Use when creating a new project. Bare AI shell by default; interactive menu adds docs / backend-infra / agents on demand.
Use after init-project to fill docs/product/start-project.md with AI research on competitors, monetization, valuation, and marketing.
Use when unsure which skill to use. Lists all vladyslav skills, workflow, and integration.
| name | attach-project |
| description | Use when adding Claude Code structure to an existing project. Auto-detects stack, creates only missing files. |
Type: Engineer (light)
Add Claude Code structure to an existing project. Never overwrites existing files. Auto-detects stacks via scripts/detect-stack.sh. The bash scaffolder (scripts/attach-project.sh) does all the work in ~0.5 seconds.
This was a Heavy Engineer skill until v3.1.0. Like init-project v3.0.0, the dispatched Sonnet subagent was doing pure mechanics (mkdir, cp, sed) — no LLM thinking needed for any of it.
Collect inputs that the scaffolder can't auto-detect.
Verify project root. Run ls -A. The scaffolder will reject the run if no project markers exist, so this is a pre-emptive sanity check. If pwd looks empty (no .git/, no language manifest like package.json / requirements.txt / etc.) → STOP and suggest /vladyslav:init-project for a new project.
Run stack detection for the user's information:
<plugin-root>/scripts/detect-stack.sh .
Parse the JSON to show the user what was detected. Example: "Detected: python, docker". This is informational — the scaffolder runs detection again internally.
AskUserQuestion — additional stacks not auto-detected (multi-select):
"Detected stacks:
<list>. Want to add any stacks not detected?" Options:python,go,flutter,swift,kotlin,other,none.
For each "other" stack — AskUserQuestion three times:
label (e.g. "Rust backend")dir (e.g. rust)gitignore entries (comma-separated, e.g. target/,Cargo.lock)AskUserQuestion — domain (free text, optional):
"Project domain / purpose? (e.g. 'iOS fitness app', 'Python data pipeline') — type
noneto skip"
AskUserQuestion — private mode (single-select):
"Private mode? Gitignore AI workflow files (CLAUDE.md, .claude/, docs/plans/)?" Options:
yes,no.
Resolve plugin root. Same approach as init-project: Glob ~/.claude/plugins/cache/vladyslav-marketplace/vladyslav/*/scripts/attach-project.sh and take the directory two levels up. Fall back to /Volumes/DevSSD/Development/vladyslav-skills (dev clone).
Execute (via the Bash tool):
<plugin-root>/scripts/attach-project.sh \
--pwd <project pwd> \
--plugin-root <plugin-root> \
[--additional-stacks "python,go,..."] \
[--other-stacks "label1:dir1:gitignore-entries1;label2:dir2:gitignore-entries2"] \
--domain "<domain-or-empty>" \
--private-mode <yes|no>
--other-stacks format: semicolon-separated list of label:dir:gitignore triples. The gitignore part is itself a comma-separated list. Pass empty string "" if no "other" stacks were chosen.
The script emits JSON:
{
"status": "success" | "error",
"detected_stacks": [<auto-detected stack names>],
"files_written": [<paths>],
"files_skipped": [<paths that already existed and were preserved>],
"warnings": [<msgs>],
"error": "<msg if status=error>"
}
Capture this output.
Parse the JSON. Render to the user:
status: success:
✓ attach-project complete
Detected: <detected_stacks joined>
Files written: <files_written count> — <list>
Files preserved (already existed): <files_skipped count> — <list, if any>
Warnings: <warnings, if any>
Next: /vladyslav:ingest — scan codebase, populate docs/, and seed MemPalace
status: error:
✗ attach-project failed
Error: <error from JSON>
The bash scaffolder, given the parameters above, writes ONLY these paths if they do not already exist:
CLAUDE.md — with auto-generated Stack section based on detected + additional stacks.gitignore — APPENDED with stack-specific entries (Python / Go / Flutter / Swift / Kotlin / Node / Web), editors/OS, secrets/env, logs, and (if private mode) AI workflow file lines. Existing entries are preserved verbatim, duplicates are skipped via grep -Fxq..claude/settings.json — with PROJECT_DOCS_ROOT env.claude/agents/docs-agent.md — documentation maintainer.claude/agents/code-review-agent.md — code-review agentdocs/architecture/system.md, docs/product/prd.md, docs/plans/tasks.md — stub files (# Title\n\n*to be filled*\n)python/, go/, swift/, flutter/, kotlin/) with .gitkeep — but ONLY if no matching directory (or sibling like backend/, app/, ios/, android/) already exists.<dir>/ with .gitkeep.Idempotent: rerun is safe. Every pre-existing file is reported in files_skipped. No file is ever overwritten.
No git init: the project already has version control. The scaffolder doesn't touch git.
No backend scaffolding: unlike init-project, this skill does NOT create requirements.txt, FastAPI main.py, Dockerfile, docker-compose files, etc. Those would conflict with the existing project. attach-project adds only the AI workflow shell.