npx claudepluginhub lklimek/agents --plugin claudiusWant just this skill?
Then install: npx claudepluginhub u/[userId]/[slug]
Unified review report format for all finding-producing agents. Load when emitting or consuming review findings.
This skill is limited to using the following tools:
Review Report Format
Unified format for all review findings. Schema: schemas/review-report.schema.json (v2.0.0).
Finding Structure
Agents emit a JSON array of finding_section objects:
[
{
"title": "Section Title",
"category": "security|project|code_quality|dependencies|documentation|pr_comments",
"findings": [
{
"id": "PREFIX-001",
"severity": 5,
"title": "Short finding title",
"tags": ["A03 Injection", "CWE-79"],
"location": "src/auth.rs:42-56",
"description": "What the issue is and why it matters",
"impact": "What could go wrong",
"recommendation": "How to fix it"
}
],
"positives": "Optional positive observations"
}
]
Required Fields
| Field | Type | Description |
|---|---|---|
id | string | PREFIX-NNN -- see ID Prefixes below |
severity | integer | 5=CRITICAL, 4=HIGH, 3=MEDIUM, 2=LOW, 1=INFO |
title | string | Short finding title |
location | string | Full file path with lines: src/auth.rs:42-56 -- never bare line numbers |
description | string | What the issue is and why it matters |
recommendation | string | How to fix it |
Optional: tags (OWASP, CWE, etc.), impact (what could go wrong).
File Output
When writing findings to a file, ALWAYS use the Write tool — never use Bash commands like cat > file, tee, heredoc redirects, or inline python3 scripts for file creation. The Write tool is allowed in all CI environments; Bash file-writing commands are typically blocked by tool allowlists.
ID Prefixes
| Prefix | Category | Used by |
|---|---|---|
SEC- | security | security-engineer-smythe |
QA- | code_quality | qa-engineer-marvin |
PROJ- | project | project-reviewer-adams |
CODE- | code_quality | developer-bilby (generic) |
RUST- | code_quality | developer-bilby (Rust) |
PY- | code_quality | developer-bilby (Python) |
GO- | code_quality | developer-bilby (Go) |
FE- | code_quality | developer-bilby (frontend) |
DOC- | documentation | technical-writer-trillian |
CMT- | pr_comments | check-pr-comments |
DEP- | dependencies | review-dependency |
IDs are provisional -- the consolidation step deduplicates and reassigns final IDs.
Domain-Specific Fields
Agents may add context to description and tags per their domain:
- security-engineer: include OWASP category and CWE in
tags, CVE references and evidence indescription - qa-engineer: include requirement reference, expected vs actual behavior in
description - check-pr-comments: include
reviewer,comment_id,comment_url,thread_id,verdictfields (schema-defined)
Report Pipeline Tools
| Tool | Purpose | Usage |
|---|---|---|
scripts/validate_report.py | Validate report JSON against schema | python3 ${CLAUDE_SKILL_DIR}/../../scripts/validate_report.py report.json |
scripts/consolidate_reports.py | Merge multiple agent reports, deduplicate findings | python3 ${CLAUDE_SKILL_DIR}/../../scripts/consolidate_reports.py agent1.json agent2.json -o consolidated.json |
scripts/generate_review_report.py | Render consolidated report as Markdown/HTML | python3 ${CLAUDE_SKILL_DIR}/../../scripts/generate_review_report.py consolidated.json |
Full Report Envelope
For complete reports (grumpy-review, check-pr-comments), wrap finding sections in:
{
"schema_version": "2.0.0",
"metadata": { "project": "...", "date": "YYYY-MM-DD" },
"executive_summary": { "overall_assessment": "..." },
"summary_statistics": { "total_findings": 0, "severity_counts": {} },
"findings": []
}
See schemas/review-report.schema.json for complete envelope schema.
Similar Skills
Search, retrieve, and install Agent Skills from the prompts.chat registry using MCP tools. Use when the user asks to find skills, browse skill catalogs, install a skill for Claude, or extend Claude's capabilities with reusable AI agent components.
Activates when the user asks about AI prompts, needs prompt templates, wants to search for prompts, or mentions prompts.chat. Use for discovering, retrieving, and improving prompts.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.