VS Code Extensions
H1VE

H1VE

by h1ve

See your current branch's feature state — stage, blockers, sign-offs, spec — from H1VE, the governance & memory layer for teams building with AI.

Downloads

3

Rating

(0)

Version

0.6.0

Last updated

Jul 20, 2026

See your current branch's feature state — right inside VS Code — from H1VE, the governance & memory layer for teams building with AI.

As your team ships with AI (Claude Code, Cursor, Copilot), context and decisions drift. H1VE keeps every feature aligned; this extension surfaces the branch's feature at a glance in the sidebar: stage, blockers, sign-offs, and the spec — no context switch to the dashboard.

Companion to the CLI h1ve and the MCP server.

h1ve.org · app.h1ve.org

What you can do (v0.3 — the extension goes active)

Beyond seeing the branch feature, the extension now lets you act without leaving the editor:

  • Command-center — click the status-bar item for a quick pick of the current stage's actions.
  • Move stage / Start-or-resume a feature / Log a blocker / Open PR — from the palette or the board.
  • Board view — every card grouped by stage, with badges ⚠ (blocker) · ⛓ (blocked by a dependency) · 📎 (attachment).
  • Copy feature context for AI — one click puts the stage, SPEC, blockers, attachments and the relevant project memory on the clipboard, ready to paste into any AI chat.
  • Recall project memory — pull only the relevant decisions/gotchas by term or selection.
  • Notifications — a toast when your card advances, CI fails, a sign-off lands, or a blocker changes.
  • Sign in — no terminal — "Sign in" runs the whole login inside the editor (loopback + browser + code exchange); the status bar shows connected as you.
  • Connect the AI agent — a button registers the H1VE MCP so Claude Code / Cursor reads your board. A Getting Started walkthrough guides the first run.
  • AI bridge (VS Code ≥ 1.95) — the Copilot agent reads your H1VE board on its own via Language Model Tools (#h1veCurrentFeature · #h1veRecall · #h1veBoard), and @h1ve answers in the native chat (/status · /recall <term> · /board). Read-only — the AI reads; it never moves cards.
  • Publish SPEC/DONE to the card — a CodeLens at the top of specs/SPEC-*.md / DONE-*.md publishes the file to the branch's card (kills the "wrote the file, forgot to publish" gap).
  • Check spec drift — surfaces anti-drift divergences (spec vs foundation) in the Problems panel, on demand.

Setup

Already using the CLI? If you've run h1ve login in a terminal, the extension uses that credential automatically — zero paste, nothing to configure. Just open the H1VE panel (the hive icon in the activity bar) on a feature branch.

Otherwise, the panel shows a "Connect to H1VE" welcome with a button. Or, manually:

  • API key: run the command "H1VE: Set API Key" — stored in the VS Code SecretStorage, never in settings. Use a PAT (nf_pat_… from app.h1ve.org/api-tokens).
  • h1ve.apiUrl (User/Machine settings — scope machine, default https://app.h1ve.org): on purpose it is not overridable by a repo's .vscode/settings.json, so a repository can't redirect where your API key is sent (OP-12).

The "Branch feature" tree refreshes every h1ve.refreshIntervalSeconds (30s default), on config change, and via "H1VE: Refresh".

Commands

Command What it does
H1VE: Set API Key Store / clear the API key in SecretStorage
H1VE: Refresh Reload the branch's feature state
H1VE: Open Spec Open the feature's spec in a markdown editor

The extension resolves the feature from your current git branch (the branch_slug recorded by h1ve start). It's read-only — to act on features (start, move, spec, done), use the CLI/MCP.


MIT · built with H1VE, by H1VE 🐝

Related extensions