groove-admin-install

Original🇺🇸 English
Translated

Install groove backends, companions, and AGENTS.md bootstrap. Run once per repo.

8installs
Added on

NPX Install

npx skill4agent add andreadellacorte/groove groove-admin-install

Tags

Translated version includes tags in frontmatter
<!-- groove:managed — do not edit; changes will be overwritten by groove update -->

groove-admin-install

Outcome

All groove backends are installed in dependency order, groove-wide companion skills are installed, AGENTS.md contains the session bootstrap, and the repo is ready for use.

Acceptance Criteria

  • Task and memory backends installed
  • Companion skills installed (find-skills, agent-browser, pdf-to-markdown)
  • AGENTS.md
    contains the
    <!-- groove:prime:start -->
    session bootstrap
  • AGENTS.md
    contains a
    <!-- groove:task:start -->
    stub (if
    tasks: beans
    )
  • User sees a summary of what was installed and what was written

Steps

Run in order:
  1. If
    .groove/index.md
    does not exist, run
    /groove-admin-config --defaults
    to create it with all defaults (no prompts)
  2. Run
    /groove-utilities-task-install
    — installs the configured task backend (e.g. beans)
  3. Run
    /groove-utilities-memory-install
    — creates memory directories
  4. Install companion skills:
    • find-skills (downloaded): check
      ls .agents/skills/find-skills/SKILL.md
      ; if absent:
      npx skills add https://github.com/vercel-labs/skills --skill find-skills
    • agent-browser (downloaded): check
      ls .agents/skills/agent-browser/SKILL.md
      ; if absent:
      npx skills add https://github.com/vercel-labs/agent-browser --skill agent-browser
    • pdf-to-markdown (embedded): check
      ls .agents/skills/pdf-to-markdown/SKILL.md
      ; if absent:
      npx skills add andreadellacorte/groove --skill pdf-to-markdown
    • Report each as installed / already-present / failed
  5. Scaffold hooks and cache directories:
    • Create
      .groove/hooks/
      if it does not exist (with a
      .gitkeep
      )
    • Create
      .groove/.cache/
      if it does not exist (with a
      .gitkeep
      )
    • Report each as created / already-present
  6. Apply git strategy — write
    .groove/.gitignore
    from
    git.*
    sub-keys in
    .groove/index.md
    (see
    /groove-admin-config
    for rules)
  7. Write the session bootstrap to
    AGENTS.md
    :
    • Replace any existing
      <!-- groove:prime:start -->
      /
      <!-- groove:prime:end -->
      section with:
      <!-- groove:prime:start -->
      **Run at the start of every session:**
      `/groove-utilities-prime` — loads groove workflow context into the conversation
      <!-- groove:prime:end -->
    • If section absent, append to end of
      AGENTS.md
      ; preserve all other content
  8. /groove-utilities-task-install
    already wrote the
    <!-- groove:task:start -->
    stub in step 2 — no additional AGENTS.md write needed here

Constraints

  • Read
    .groove/index.md
    for
    tasks:
    and
    git.*
    config before running
  • If
    .groove/index.md
    does not exist,
    /groove-admin-config
    is run first (step 1) to create it
  • Dependency order for backends must be respected: task → memory → companions
  • Each step reports installed / already-present / failed
  • AGENTS.md
    update is additive per section — preserve all other content
  • If any step fails, report it clearly but continue with remaining steps
  • Companion skills (find-skills, agent-browser, pdf-to-markdown) are hardcoded here, not read from any config file
  • Report a final summary:
    ✓ task backend (beans)
    ✓ memory backend — memory dirs ready
    ✓ companion: find-skills
    ✓ companion: agent-browser
    ✓ companion: pdf-to-markdown
    ✓ hooks: .groove/hooks/ ready
    ✓ AGENTS.md updated (groove:prime, groove:task)