spacedock
Staged workflows with human approval gates and recorded decisions
All Skills in This Repository (11)
Pure Emerald Level Indicatorsdebrief
Automates creation of debriefs from workflow sessions, capturing commits, decisions, and issues.
commission
Design, generate, and launch plain text workflows with stages and entities.
first-officer
Orchestrate Spacedock multi-agent workflows with approval gates and entity state advancement.
present-gate
Renders FO gate decisions into standardized captain-facing presentations with assembly rules and review findings.
using-legacy-claude-team
Manage legacy Claude Team creation, registry recovery, and teardown via TeamCreate.
refit
Upgrade workflows to the latest Spacedock version with schema migration.
fo-status-viewer
Query and update Spacedock workflow and entity statuses.
survey
Analyze AI-agent session history to identify implicit workflows and open decisions.
ensign
Automate multi-stage workflow execution and decision tracking across AI agents.
feedback-rejection-flow
Route rejected feedback, track cycles, and escalate in multi-agent workflows.
fo-dispatch-recovery
Automate Claude dispatch recovery for degraded modes and context budget failures.
Frequently Asked Questions
FAQPage SchemaHow to install Spacedock?โผ
Run `npx skills add spacedock-dev/spacedock --all -g -y` in your terminal to install all skills in this suite globally.
What problem does Spacedock solve?โผ
It stops agents from judging their own work by routing every stage through an approval gate with recorded evidence, so you decide on facts instead of transcripts.
Does Spacedock work with Claude Code and Codex?โผ
Yes. It runs inside Claude Code, Codex, and Pi as tier-1 hosts, and works in most other harnesses through standard skill systems.
Can Spacedock analyze an existing project?โผ
Yes. The survey skill reads your past agent session history and reports the workflow you have been running, open decisions, and how often you had to step in.
Do I need to write code to use Spacedock?โผ
No. You design workflows by answering a few plain-English questions, and the agent generates all workflow files and runs the stages for you.