Team-level knowledge layer for neuroflow. Connects a researcher's personal neuroflow project to a shared GitHub organisation repo where team research directions, cross-project findings, and recommended methods are coordinated. Use when the user wants to sync their work with a team, share findings explicitly, view team directions, or get team-aware recommendations. Never automatically copies personal project data to Hive — all sharing is explicit.
From neuroflownpx claudepluginhub stanislavjiricek/neuroflow --plugin neuroflowThis skill uses the workspace's default tool permissions.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Executes pre-written implementation plans: critically reviews, follows bite-sized steps exactly, runs verifications, tracks progress with checkpoints, uses git worktrees, stops on blockers.
Guides idea refinement into designs: explores context, asks questions one-by-one, proposes approaches, presents sections for approval, writes/review specs before coding.
Hive is the team-level counterpart to flowie. Where flowie tracks an individual researcher's project workflow and session state, Hive connects to a shared GitHub organisation private repo that the whole lab or team uses to:
Privacy rule (enforced absolutely): Nothing from a personal .neuroflow/ project is ever automatically sent to Hive. Every share is an explicit, intentional action by the researcher. The Hive is a shared workspace, not a surveillance layer.
{org}/{hive-repo}/
├── hive.md ← team identity: who we are, what we study, norms
├── directions.md ← active research directions (updated by PIs / team leads)
├── sync.json ← sync metadata: last pull per member, last push timestamps
└── shared/
├── methods/ ← recommended analysis methods and pipelines
├── literature/ ← curated papers the team wants everyone to read
└── findings/ ← explicitly shared results and summaries
When a project has joined a Hive, the following folder exists:
.neuroflow/hive/
├── hive.md ← local copy of team identity (pulled from org repo)
├── directions.md ← local copy of team research directions
└── sync.json ← sync metadata: hive_repo URL, last_pull, last_push, member_handle
This skill is invoked by the /hive command, which supports five modes:
--initConnect the current neuroflow project to a Hive repo for the first time.
{org}/{hive-repo}member_handle in sync.json)hive.md and directions.md.neuroflow/hive/ with local copies of both files and an initial sync.json.neuroflow/flow.md to add a hive/ rowhive_repo: {org}/{hive-repo} and hive_member: {handle} to project_config.mdhive.md and the active directions from directions.md--syncPull the latest state from the Hive repo and update local copies.
hive.md, directions.md, and sync.json from the org repo.neuroflow/hive/hive.md and directions.mdsync.json with last_pull: [timestamp]--viewDisplay the current state of the team Hive without syncing.
.neuroflow/hive/hive.md and directions.mdsync.json to show when last synced/hive --sync to fetch the latest updates from the team."--shareExplicitly share a finding, method, or curation from this project to the Hive.
This is the only way anything from a personal project reaches the Hive. It is always user-initiated.
shared/{category}/{slug}.md in the Hive repo via GitHub API (or gh CLI if available)sync.json with last_push: [timestamp] and a log entryshared/{category}/{slug}.md"If the user has not connected to a Hive (--init not run), stop and prompt them to run /hive --init first.
--recommendGet team-aware recommendations for the current project phase.
hive.md and directions.mdproject_config.md to know phase, modality, and research questionshared/methods/ and shared/literature/ in the Hive (if accessible) for relevant shared content| What | Shared to Hive? |
|---|---|
| Research question | Never automatically — only if user explicitly runs --share |
| Session logs | Never |
| Raw data paths or outputs | Never |
| Analysis results | Never automatically — only with --share |
| Personal project_config.md fields | Never |
Something the user explicitly approves via --share | Yes, after confirmation |
The Hive is pull-first: the researcher benefits from team knowledge without being required to share anything back.
Hive uses the same GitHub credentials as the user's local git config. The gh CLI (GitHub CLI) is preferred for push operations — check with gh auth status. If not available, fall back to constructing GitHub API calls with a PAT (personal access token) that the user provides.
Authentication instructions:
gh auth login # recommended
# or configure git credentials for HTTPS push
neuroflow:neuroflow-core — read first; defines the command lifecycle and .neuroflow/ write rulesneuroflow:phase-flowie — the personal-project counterpart to Hive; understand flowie before implementing Hive to avoid overlap