init-specslisted
Install: claude install-skill datamaker-kr/synapse-claude-marketplace
# Init Specs
## Skill Info
Part of the **spec-manager** agent. This skill scaffolds the initial spec documents for a new task.
The pipeline that `init-specs` scaffolds depends on the task's **difficulty**:
- `low` → **lite pipeline**: only `requirements.md` and `plans.md` are created. `specs.md` is skipped. The next step is `/plan-with-requirements`.
- `medium` / `high` → **full pipeline**: `requirements.md`, `specs.md`, `plans.md` are all created (legacy behavior). The next step is `/specify-with-requirements`.
If Jira MCP is available **and** a ticket ID is provided, `init-specs` will also auto-populate `requirements.md` from the Jira issue.
## Input
The user provides a task title as arguments: $ARGUMENTS
Arguments may optionally include a ticket ID prefix and the flags below:
- `/init-specs SYN-1234 User Authentication Flow` — ticket ID + title
- `/init-specs User Authentication Flow` — title only (ticket ID asked later)
- `/init-specs SYN-1234 Fix copy typo --difficulty low` — force lite pipeline
- `/init-specs SYN-1234 Big refactor --difficulty high --pipeline full` — force full pipeline
- `/init-specs SYN-1234 Some task --no-jira` — skip Jira auto-fill even if MCP is available
Recognized flags (any order, before or after the title):
| Flag | Values | Default | Effect |
|------|--------|---------|--------|
| `--difficulty` | `low` / `medium` / `high` | inferred | Forces difficulty. Skips inference. |
| `--pipeline` | `lite` / `full` | derived from difficulty |