Quickstart

1. Install the CLI

pip install stndp-cli

This gives you the stn command (requires Python 3.11+).

2. Log in

stn login

This opens your browser to sign in. When you finish, the CLI stores a token locally (owner-readable only) and you're authenticated. Check who you are any time:

stn whoami

3. Get a team

Entries live in a team. Create one you own:

stn team create "Hydra Engineering"

…or, if a teammate invited you, list your teams and accept any pending invite:

stn team list
stn team invites
stn team accept TEAM_ID

Each team has a slug (for example hydra-engineering) you pass to the commands below — the team's name works too. To avoid repeating --team, make one your default:

stn team use hydra-engineering   # team-scoped commands now use it unless you pass --team

4. Post your first update

stn push hydra-engineering "Done: shipped auth refactor. Doing: payments. Blocked: need @sarah for integration tests."

@mention teammates inline — every @handle must be a real user or the push is rejected. To flag a blocker explicitly, add --blocked; set its urgency with --severity high (high and critical relay to the team immediately).

5. Read the feed

stn team feed hydra-engineering   # the whole team, with each block's resolutions nested beneath it
stn status hydra-engineering      # just today
stn log hydra-engineering         # your own recent entries (last 7 days)
stn inbox --unresolved            # blocks that @mention you and that you haven't cleared yet

6. Save your working state

Before you switch tasks or stop for the day, capture what you're in the middle of — privately, with your local git context attached automatically:

stn checkpoint "refactoring the webhook retry path" --next "wire up the dead-letter queue"

Next time you sit down, reload it — along with any open blockers and the PRs awaiting your review:

stn resume

7. Wire up your coding agent (optional)

Let Claude Code, Cursor, Codex, Gemini CLI, or Windsurf read and write the same shared context over MCP:

stn agent setup     # writes the MCP config + a short rules block; auto-detects your agents
stn agent doctor    # confirm the wiring (login, MCP server, config, rules, hook)

This writes one line of MCP config plus a short rules block into your project; restart your agent and it can resume, search, decide, and checkpoint as it works. The rules block points at stn guide — the live workflow playbook an agent (or you) can print any time, with stn <command> --help for the details of any command. See Coding agents (MCP).

8. Connect your tools (optional)

Connect GitHub, Jira, and Slack so your work links itself into the graph — no manual tagging. Connect each from your dashboard under Integrations (read-only OAuth; tokens are encrypted at rest). What you get:

See Integrations for setup details and exactly what each connection reads.

9. Your data, anytime

stn export hydra-engineering     # your entries as Markdown (./stndp-export)
stn export --me                  # your private checkpoints instead

That's it. For the complete command set — decisions, search, the context graph, schedules, API tokens, and configuration — see the CLI reference and How stndp works.