Skip to content
Back to skills

Review Decision

ASecurity

Review a decision record for clarity, completeness, reasoning quality, and consequence coverage.

  • 10 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 6, 2026
databasesgosqldatabase

Security analysis

A100/100

Scanned October 6, 2026

npx -y skills add tomzx/agents --skill review-decision --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Review Decision?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Review Decision
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tomzx-review-decision/badge)](https://www.skillsdirectory.com/skills/tomzx-review-decision)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: review-decision
description: Review a decision record for clarity, completeness, reasoning quality, and consequence coverage.
argument-hint: "[decision document or file path]"
---

# Review Decision

Audits a decision record and reports findings across four categories: clarity, completeness, reasoning quality, and consequence coverage.

## Prerequisites

- Apply the shared SDLC conventions in `skills/sdlc/references/shared.md`.
- `.sdlc/knowledge/decisions/N-<slug>.md`, or a decision document provided in context or as a file path

## Steps

1. Read the decision document.
2. Evaluate it against the checklist below.
3. Report findings by category. Omit categories with no findings.
4. After the review, update `status` in frontmatter to the appropriate value: `Accepted`, `Deprecated`, or `Superseded by [N]`.

## Review Checklist

### Clarity
- Is the decision stated as a specific, unambiguous choice (not a direction or intent)?
- Is the title a noun phrase that names the decision, not just the topic?
- Is the status field present and accurate (Proposed / Accepted / Deprecated / Superseded)?

### Completeness
- Is the context section present and does it explain what problem or question prompted the decision?
- Are at least two options described, including the one chosen?
- Does each option have at least one pro and one con?
- Are trade-offs (gained vs. sacrificed) explicitly stated?

### Reasoning Quality
- Does the reasoning connect the context and constraints to the chosen option?
- Are pros/cons specific to this situation rather than generic statements?
- Is it clear why the chosen option was preferred over the alternatives?
- Are any options dismissed without justification?

### Consequence Coverage
- Are the practical effects on the codebase or process described?
- Are follow-up decisions or deferred work identified?
- Are risks introduced by this decision acknowledged?

## Output Format

```markdown
## Clarity

<Findings or "No issues found.">

## Completeness

<Findings or "No issues found.">

## Reasoning Quality

<Findings or "No issues found.">

## Consequence Coverage

<Findings or "No issues found.">
```

## Outcome

If `$OUTCOME_YAML` is set, emit your verdict there per `skills/sdlc/references/shared.md`:

| Verdict | When |
|---|---|
| `approved` | No blocking findings; the subject passes review |
| `changes-requested` | Findings the author must address before it passes |
| `rejected` | Fundamental flaw requiring rework or stopping |

In the same emission, list the record under `artifacts:` when the review changed it (for example a status transition or a recorded answer).

## Example Usage

**Scenario 1: Vague decision statement**
Decision title is "Database approach" and the decision section says "we'll use PostgreSQL where appropriate."
Report under Clarity: the decision must state specifically what was chosen and for what purpose.

**Scenario 2: Single option considered**
Only the chosen approach is listed; no alternatives are documented.
Report under Completeness: at least one alternative must be described so future readers understand what was ruled out.

**Scenario 3: Unjustified dismissal**
Option B is listed but has no pros and only one con, with no explanation of why it was rejected.
Report under Reasoning Quality.

**Scenario 4: No follow-up captured**
A decision to use a flat-file config acknowledges it won't scale but records no follow-up task to revisit the choice.
Report under Consequence Coverage.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…