2.6 KiB
2.6 KiB
Code Review
Thorough code review with test verification and structured feedback.
Usage
/cipherpowers:code-review [--model=<sonnet|opus|haiku>]
Model guidance:
opus- Deep analysis, security-critical code, complex architecturesonnet- Balanced quality/speed (default if not specified)haiku- Quick reviews, simple changes
MANDATORY: Skill Activation
Load skill context: @${CLAUDE_PLUGIN_ROOT}skills/conducting-code-review/SKILL.md
Step 1 - EVALUATE: State YES/NO for skill activation:
- Skill: "cipherpowers:conducting-code-review"
- Applies to this task: YES/NO (reason)
Step 2 - ACTIVATE: If YES, use Skill tool NOW:
Skill(skill: "cipherpowers:conducting-code-review")
⚠️ Do NOT proceed without completing skill evaluation and activation.
Algorithmic Dispatch
Decision tree (follow exactly, no interpretation):
-
Is this a code review request?
- YES → Continue to step 2
- NO → This command was invoked incorrectly
-
Have you already dispatched to code-review-agent agent?
- YES → Wait for agent to complete
- NO → Continue to step 3
-
DISPATCH TO AGENT NOW:
Use Task tool with:
subagent_type: "cipherpowers:code-review-agent"
model: [from --model arg if provided, otherwise omit to use default]
description: "Code review workflow"
prompt: """
[User's original request or task context]
Follow the conducting-code-review skill exactly as written.
Review the recent changes and provide structured feedback.
"""
Model parameter rules:
- If user specified
--model=X→ passmodel: Xto Task tool - If no model specified → omit model parameter (agent default applies)
- STOP. Do not proceed in main context.
Why Algorithmic Dispatch?
- 100% reliability: No interpretation, no rationalization
- Agent enforcement: Persuasion principles prevent rubber-stamping
- Consistent quality: Every review runs tests, checks all severity levels
- Skill integration: Agent reads conducting-code-review skill automatically
What the Agent Does
The code-review-agent agent implements:
- Identify code to review (git commands)
- Review against practice standards (ALL severity levels)
- Save structured feedback to work directory
- No approval without thorough review
Note: Tests and checks are assumed to pass. The reviewer focuses on code quality, not test execution.
References:
- Agent:
${CLAUDE_PLUGIN_ROOT}agents/code-review-agent.md - Skill:
${CLAUDE_PLUGIN_ROOT}skills/conducting-code-review/SKILL.md - Standards:
${CLAUDE_PLUGIN_ROOT}standards/code-review.md