Slash commands

The /ritual skill registers four subcommands inside any supported coding agent. That’s the entire surface — there are no other slash commands.

Quick reference

Command	What it does	Time
/ritual build	The full context → scope → discovery → recommendations → build brief → implementation loop	20–60 min
/ritual resume	Pick up an in-flight or paused exploration where you left off	5–10 sec
/ritual lineage	Reverse lookup: which explorations + decisions shaped this file?	5–10 sec
/ritual context-pulse	Score Reasoning Readiness / Context Debt before code is written	10–30 sec

/ritual build

The main pipeline. Takes a one-line feature description and walks through a structured loop — context → scope → discovery → recommendations → build brief → implementation — with the agent pausing where your judgment is needed.

/ritual build Add an AI assistant to the product

The loop runs through six named stages (same names as the CLI’s progress header — ● Context ○ Scope ○ Discovery ○ Recommendations ○ Build brief ○ Implementation):

Stage	What happens
Context	Scan the codebase, map candidate files, surface prior decisions from the workspace KG
Scope	Generate sub-problems, pick which to solve for, lock the developer-oriented problem frame
Discovery	Generate classified questions (🔴 blocking / 🟡 uncertain / 🟢 answerable). You pick + answer the ones that matter.
Recommendations	Agentic pipeline generates structured options across the open decisions (model, data scope, integration boundaries, etc.); you accept or refine each one, locking an approved snapshot
Build brief	Implementation-ready delivery context written to BUILD-BRIEF.md — rationale, rejected alternatives, acceptance criteria
Implementation	Coding agent codes from the brief. Review-blocking items first. KG writebacks happen here as code lands — no separate “sync” step.

See Build stages for the deep reference on each stage.

Use when: you have a feature idea and want Ritual to handle structured discovery + decisions + implementation end-to-end.

/ritual resume

Restores the most recent (or named) exploration into the agent’s working context.

/ritual resume
/ritual resume add-ai-assistant

Reloads:

• Answered discovery questions
• Accepted recommendations
• The Build Brief
• Current stage + last completed step

Useful when you’ve switched machines, returned after a break, or are jumping into a build started in another agent session.

Use when: you started a build, walked away, and want to pick up exactly where you stopped without re-explaining context.

/ritual lineage

Reverse lookup: paste a file path (or several) and get back every prior exploration, decision, or deferral that touched those files.

/ritual lineage src/auth/permissions.ts
/ritual lineage src/auth/ src/middleware/auth.ts

Returns:

• Explorations whose recommendations led to changes in these files
• Decisions captured during those explorations
• Open deferrals that may collide with new work here

Use when: you’re about to edit a file and want to know what shaped its current shape — before you assume.

/ritual context-pulse

Scores Reasoning Readiness (and the inverse, Context Debt) for a feature description or an existing exploration. The pulse names what’s missing — open questions, unverified assumptions, weakly-grounded codebase claims — before code is written.

/ritual context-pulse Add an AI assistant to the product
/ritual context-pulse exp-7a2b9c
/ritual context-pulse                        # naked mode — agent asks what you want to pulse

Output (example):

  Reasoning Readiness    18 %
  Context Debt           82 %
  Context Surface        Raw ask

  Feature clarity        8 / 30
  Decision resolution    0 / 30
  Repo grounding         10 / 25
  Assumption safety      0 / 15

  Top debt sources:
    • No acceptance criteria — what does "done" look like?
    • 10 unresolved decisions across model, RAG, memory, UI, eval
    • 3 unverified assumptions about existing infra

  Next: /ritual build "Add an AI assistant to the product"

The pulse also fires inline during /ritual build — at the end of each major stage you see the readiness number climb as the agent gets grounded.

See Context Pulse for the deep reference on dimensions, modes, and scoring.

Use when: you want to know if a feature ask is ready for /ritual build, or to score what changed since the last build.

Why only four commands?

Earlier versions of Ritual shipped many slash commands (/ritual explore, /ritual run, /ritual brief, /ritual gate, etc.). Each was a thin wrapper around an MCP tool call, which the agent can do on its own from plain English. The four commands above are the ones that need a slash-command surface because they kick off long-running, multi-stage flows that benefit from a named entry point.

Anything you’d want to do that isn’t one of these four (e.g. “list my recommendations,” “check the status of an exploration”) — just ask your agent in plain English. The MCP tools handle it.

Related

• CLI — install + setup
• Build stages — deep dive on the six stages inside /ritual build
• Context Pulse — Reasoning Readiness + Context Debt scoring
• MCP server — direct integration over Streamable HTTP