---
id: "/patterns/"
title: "Misalignment is mostly a skill issue."
creator: null
url: "https://skillissue.dev/patterns/"
level: null
category: null
models: null
source_date: null
date_added: null
---

# Misalignment is mostly a skill issue.

This is a free library of prompt and spec patterns for people who build with AI agents.
When an agent does the wrong thing, the fix is usually a clearer spec and a verification gate,
not a different model. The harness you put around the model matters more than the model itself.

Every pattern has the problem it solves, how it works, a copy-paste template, and when not to use it.

## Patterns

All
Specs (2)
Verification gates (3)
Agent harness (3)
Context (2)
Prompting basics (2)

12 patterns

- [Specs

### Spec-first task brief

Write the goal, inputs, outputs and done-criteria down before the agent touches anything.
](/patterns/spec-first-brief.html.md)

- [Specs

### Acceptance criteria checklist

Turn "done" into a list of observable, yes-or-no checks the agent and you can both verify.
](/patterns/acceptance-criteria.html.md)

- [Verification gates

### Tests as the gate

Define the check first, then let the agent iterate until the check passes. Not until it feels done.
](/patterns/tests-as-gate.html.md)

- [Verification gates

### Evidence-of-done report

Require the agent to end with what it ran, what it saw, and what it did not verify.
](/patterns/evidence-of-done.html.md)

- [Verification gates

### Fresh-eyes review pass

Review the diff in a new session that sees only the spec and the change, not the reasoning that produced it.
](/patterns/fresh-eyes-review.html.md)

- [Agent harness

### Plan, checkpoint, then execute

Have the agent propose a plan and stop. Approve or correct it before any changes are made.
](/patterns/plan-then-execute.html.md)

- [Agent harness

### Approval gate for irreversible actions

Let the agent draft anything it likes, but require a human yes before it sends, deletes, pays or publishes.
](/patterns/approval-gate.html.md)

- [Agent harness

### One task per run, small diffs

Break work into small, separately verifiable tasks and commit between them.
](/patterns/small-batches.html.md)

- [Context

### Project rules file

Keep stable project knowledge in one versioned file the agent reads every session.
](/patterns/project-rules-file.html.md)

- [Context

### Handoff summary between sessions

End long sessions with a compact state file, and start the next session from it instead of the full history.
](/patterns/handoff-summary.html.md)

- [Prompting basics

### Delimit instructions from data

Wrap pasted documents, logs and user content in clear tags so the model can tell them apart from your instructions.
](/patterns/delimit-inputs.html.md)

- [Prompting basics

### Few-shot examples

Show two to five worked input/output examples instead of describing the format in prose.
](/patterns/few-shot-examples.html.md)