- Home
- Skills
- Openharmonyinsight
- Openharmony Skills
- Cpp Core Guidelines Review
cpp-core-guidelines-review_skill
- Python
0
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openharmonyinsight/openharmony-skills --skill cpp-core-guidelines-review- SKILL.md9.2 KB
Overview
This skill runs a parallel, sectioned review of C++ code against the C++ Core Guidelines using specialized sub-agents. It coordinates one sub-agent per guideline section to produce per-section markdown reports in review/ and then consolidates those results into a Summary.md. Use it to find high-confidence guideline violations across files or directories without modifying source code.
How this skill works
I first clarify the exact scope (target files, patterns, exclusions). Then I launch dedicated sub-agents in parallel, one per guideline section, each configured to read its section file and analyze only the specified target files. Each sub-agent writes a <SectionName>.md with file path, line numbers, code snippets, and specific rule IDs for confirmed violations. After all agents complete, I aggregate their outputs into review/Summary.md with counts, rankings, and recommendations.
When to use it
- Review specific C++ files or a directory for nonconforming patterns against the Core Guidelines
- Perform a focused review limited to certain guideline sections (e.g., Resource Management)
- Generate a consolidated compliance report for audits or code quality gates
- Identify systemic patterns and high-priority violations across a codebase
- Automate recurring guideline checks without changing source files
Best practices
- Always provide explicit scope: file paths, extensions, and exclusions before starting
- Limit the review to relevant sections when you want faster, targeted results
- Keep the repository's references/Content/ updated so each sub-agent has the correct ruleset
- Allocate sufficient timeouts for large codebases; the process waits for each sub-agent to finish
- Review the generated per-section markdown files before applying fixes
Example use cases
- Run a full C++ Core Guidelines audit for a new module before release
- Check only Resource Management and Error Handling sections for memory/safety issues
- Scan all header files in include/ for interface and const-correctness violations
- Produce a summarized report for a security review showing high-priority guideline breaches
FAQ
No. The process only writes to the review/ directory. It never modifies reviewed source files.
Can I limit the review to specific guideline sections?
Yes. You can request a subset of sections; the skill launches one sub-agent per requested section.
What does each section report contain?
Each <SectionName>.md lists confirmed violations with file path, line numbers, code snippet, and the exact rule ID(s) violated.