⚒ ACG ⚒

AI Rituals

Repeatable prompting and review habits Guild members use to represent working knowledge and keep AI work grounded, inspectable, and useful.

This page begins the Guild's collection of rituals for working with AI. A ritual is not a magic phrase and not a superstition. It is a repeatable habit that shapes how a human approaches an AI system before, during, and after prompting.

The Guild cares deeply about knowledge representation: how working knowledge gets stored, transmitted, tested, and reused. One way to represent that knowledge is through rituals. A good ritual compresses experience into a form that another craftsperson can repeat without losing the important structure.

The point is not to sound clever. The point is to reduce drift, surface assumptions, improve verification, and keep human judgement active. Prompting without ritual easily turns into vague delegation. Prompting with ritual keeps the craftsperson in the loop.

Here are the first six rituals in the archive. The initial seed set came from Alex Bunardzic; the archive will grow through member contribution.

Starter Rituals

The first ten rituals in the Guild archive. This list is expected to grow as members contribute their working practices.

Ritual 01

Name The Job

Start by telling the model exactly what role it is playing and what job it is doing right now. Not "help me with this." Name the task boundary clearly.

This reduces fuzzy assistance and improves the odds of getting an artifact you can inspect instead of a cloud of adjacent words.

Prompt pattern You are helping me as a reviewer / debugger / spec editor. The job is: [one concrete job]. Do only that job.
Ritual 02

Declare Non-Negotiables First

State the constraints before you ask for ideas: architecture boundaries, safety constraints, style limits, files that must not change, evidence requirements, business rules.

AI tends to fill empty space. Constraints should arrive before generation, not after cleanup.

Prompt pattern Constraints: - do not change X - stay within Y - preserve Z Now propose the smallest valid approach.
Ritual 03

Ask For The Failure Mode

Before accepting the happy path, ask the model how its own answer could fail. This shifts the interaction from affirmation to examination.

Guild work improves when the model is used to surface breakpoints, missing assumptions, and likely regressions before implementation.

Prompt pattern Before you continue, tell me the top 3 ways this answer could be wrong, incomplete, or risky.
Ritual 04

Separate Generation From Judgment

Use one pass to generate options and a second pass to evaluate them. Do not ask the model to invent and certify in the same breath.

This is the prompting equivalent of separating implementation from review.

Prompt pattern Pass 1: generate 3 options. Pass 2: critique each option against these criteria. Do not merge the two passes.
Ritual 05

Force Explicit Assumptions

If the model is making assumptions, make it say them out loud. Hidden assumptions are where most prompt drift begins.

Once assumptions are visible, you can confirm, reject, or narrow them before they contaminate the rest of the exchange.

Prompt pattern List the assumptions you are making. Mark each as confirmed, inferred, or unknown.
Ritual 06

End With Verification

Every meaningful prompting sequence should end with a verification question: what changed, how do we test it, what would falsify it, what still needs human review.

The ritual is complete only when the output is tied back to reality.

Prompt pattern Summarize: 1. what changed 2. how to verify it 3. what remains uncertain 4. what a human must still check

How To Add A Ritual

The archive should grow by contribution, not by random accumulation.

A Guild ritual should be specific enough to repeat and concrete enough to test. If a ritual cannot be demonstrated in practice, it does not belong here yet.

In that sense, a ritual is a knowledge artifact. It captures not just what to type, but how to think, what to check, and where failure usually hides.

Suggested contribution format:

  • Name: short, memorable, operational.
  • Intent: what failure mode does it reduce?
  • Trigger: when do you use it?
  • Pattern: the actual wording or sequence.
  • Evidence: what improvement have you observed?
  • Failure mode: how can the ritual itself be misused?

Over time this page can evolve from a simple list of prompting habits into a real Guild practice manual for human-AI collaboration.

Continue The Archive

Use this page as the beginning of a shared Guild vocabulary for prompting discipline, review habits, and human-in-the-loop AI practice.