groove-utilities-onboard
SKILL.md
groove-utilities-onboard
Generate a GROOVE.md file at the project root — a concise onboarding guide for contributors joining a project that uses groove. Also optionally appends a groove section to CONTRIBUTING.md if it exists.
Outcome
New contributors understand how groove works in this project: which commands to run, where logs live, and what the compound loop expects of them.
Acceptance Criteria
GROOVE.mdexists at the project root with project-specific config values substituted- If
CONTRIBUTING.mdexists, it contains a "## Using groove" section linking toGROOVE.md - Content reflects actual configured values (task backend, git strategy)
Steps
- Read
tasks.backend,memory.review_days,git.*from.groove/index.md; memory is always.groove/memory/, specs is always.groove/memory/specs/ - Ask: "Any project-specific context to include? (e.g. which tasks to pick up, team conventions) — enter to skip"
- Write
GROOVE.mdfrom the template below, substituting live config values - If
CONTRIBUTING.mdexists and does not already contain<!-- groove:onboard -->:- Append the following section:
<!-- groove:onboard --> ## Using groove This project uses [groove](https://github.com/andreadellacorte/groove) for AI-assisted engineering workflow. See [GROOVE.md](./GROOVE.md) for setup and commands. <!-- groove:onboard:end -->
- Append the following section:
- Report: "GROOVE.md written" and whether CONTRIBUTING.md was updated
GROOVE.md template
# Groove Workflow
This project uses [groove](https://github.com/andreadellacorte/groove) — an AI-assisted engineering workflow companion.
## Quick start
```bash
npx skills add andreadellacorte/groove
/groove-admin-install
Then start every session with:
/groove-utilities-prime
Daily workflow
/groove-daily-start — review yesterday, load tasks, set agenda
/groove-daily-end — write memory, commit logs
Compound loop (for any feature or fix)
/groove-work-brainstorm — clarify scope
/groove-work-plan — research and plan
/groove-work-exec — implement
/groove-work-review — evaluate output
/groove-work-compound — capture lessons
Config
| Key | Value |
|---|---|
| Task backend | <tasks.backend> |
| Memory path | .groove/memory/ |
| Specs path | .groove/memory/specs/ |
| Git strategy (memory) | <git.memory> |
Conventions
- Stage tasks are named
YYYY-MM-DD, <Stage>— no numbers - Task completion requires a "Summary of Changes" in the task body before marking done
- Archive is always user-triggered — never automatic
Where things live
- Config:
.groove/index.md - Daily logs:
.groove/memory/daily/ - Learned insights:
.groove/memory/learned/ - Specs:
.groove/memory/specs/ - Hooks:
.groove/hooks/(start.md, end.md)
Constraints
- Read
.groove/index.mdand substitute all placeholder values before writing - If
GROOVE.mdalready exists: ask "GROOVE.md exists — overwrite, update, or skip?" before touching it - The
<custom-context>block is replaced with user-provided context or removed if none - Do not create
CONTRIBUTING.mdif it doesn't exist — only append to existing files GROOVE.mdshould be short enough to read in 2 minutes
Weekly Installs
32
Repository
andreadellacorte/grooveGitHub Stars
4
First Seen
7 days ago
Security Audits
Installed on
opencode32
gemini-cli32
github-copilot32
codex32
kimi-cli32
cursor32