Overview of the orcy hierarchical model — Mission -> Task -> Subtask, status rules, authentication, and tool selection
Scanned 9/3/2026
Install to Claude Code
npx -y skills add waterworkshq/orcy --skill orcy-overview --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Orcy Overview?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/waterworkshq-orcy-overview)More formats (shields.io, HTML) on the badges page.
---
name: orcy-overview
description: Overview of the orcy hierarchical model — Mission -> Task -> Subtask, status rules, authentication, and tool selection
license: MIT
---
# Orcy Overview — Hierarchical Model & Authentication
This skill defines the **data model, authentication, and decision framework** for Orcy. It is a companion to the CLI usage and MCP usage skills.
---
## Hierarchy
```
Habitat
+-- Mission (orcy card, flows through columns)
+-- Task (work unit, claimed by agents)
+-- Subtask (simple checklist item)
```
### Habitat
The top-level container. A habitat represents a project or sprint with configurable columns (e.g., Backlog, In Progress, Review, Done). Habitats contain missions.
### Mission
A mission is a kanban card that moves across columns. It represents a product initiative (e.g., "Implement Authentication"). Missions are **not** assigned to agents — they contain tasks which are.
Mission properties:
- Title, description, acceptance criteria
- Priority (low, medium, high, critical)
- Labels for categorization
- Dependencies on other missions (`dependsOn` / `blocks`)
- SLA timeout, due date
- Status: **auto-derived** from child task states
### Task
A task is a unit of work inside a mission. Agents **claim** tasks to lock them, then work through the lifecycle.
Task properties:
- Title, description, priority
- `requiredDomain` (frontend, backend, devops, testing, fullstack)
- `requiredCapabilities` (e.g., `["typescript", "postgresql"]`)
- `estimatedMinutes`, `version` (optimistic locking)
- Dependencies on other tasks
- Quality checklist with gated completion
### Subtask
Simple checklist item within a task. Has a title, completion flag, and optional assignee. No lifecycle — just done/not-done.
---
## Mission Status — Auto-Derivation
Mission status is **never set manually**. It is computed from child task states:
| Mission Status | Condition |
|---------------|-----------|
| `not_started` | All tasks are pending |
| `in_progress` | Any task is claimed, in_progress, submitted, approved, or rejected |
| `review` | All tasks submitted, approved, or done; none pending/in_progress/claimed |
| `done` | All tasks done or approved; at least one is done |
| `failed` | Any task failed and none being actively worked on |
### Column Mapping
Missions auto-advance to the correct column based on derived status:
| Status | Target Column |
|--------|--------------|
| `not_started` | First column (Backlog) |
| `in_progress` | Second column (In Progress) |
| `review` | Second-to-last non-terminal column (Review) |
| `done` | Terminal column (Done) |
| `failed` | Stays in current column |
---
## Artifact Types
When submitting artifacts (PR links, logs, screenshots), use one of these types:
| Type | When to Use |
|------|-------------|
| `pr` | Pull request URL (most common for code tasks) |
| `commit` | Direct commit link |
| `file` | Link to a modified file |
| `screenshot` | Visual evidence of changes |
| `log` | Build output, test results, error logs |
---
## Authentication & Environment Variables
Set these in your environment or `.env`:
```
ORCY_API_URL=http://localhost:3000
ORCY_HABITAT_ID=your-habitat-uuid
ORCY_AGENT_ID=your-agent-uuid
ORCY_API_KEY=your-api-key
```
The API key authenticates all requests. The agent identity is **derived from the key** — request body fields for agent ID are ignored by the server. This prevents impersonation.
For agent registration (first time), set `ORCY_REGISTRATION_TOKEN` if your server requires one.
---
## When to Use CLI vs MCP
| Scenario | Recommended Tool | Reason |
|----------|-----------------|--------|
| Interactive shell session | CLI (`orcy ...`) | Faster for ad-hoc queries |
| Agent automation in code | MCP (`orcy_habitat({...})`) | Structured input/output, no parsing |
| Quick habitat summary | CLI `orcy habitat summary` | Single command, grep-friendly |
| Within an agent session | MCP | Tool calls are native to agent context |
| Daemon management | CLI `orcy serve start --detach` | Only available via CLI |
| Bulk operations | CLI `orcy admin batch-*` | Shell piping, scripting |
| Cross-agent communication | MCP (`orcy_habitat_message({...})`) | Agent-readable structured messages |
### Prefer MCP for Intra-Session Tool Use
If you have the MCP server configured, use MCP dispatch tools when operating from within an agent session. The CLI is best for human-driven shell workflows, debugging, and daemon management.
---
## Critical: Context Before Action
Always call `orcy_habitat({ action: "summary" })` or `orcy habitat summary` FIRST. Understand the habitat state before diving into individual missions. This prevents context pollution from loading every mission individually.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!