Claude-skill-registry kantra-analyze

Utilize the command

/kantra-analyze

to perform analysis on application source code for violations using Kantra. Be sure to ask the user what directory to analyze, as well as suggest target technologies.

IMPORTANT: If the user makes changes to what you suggest, be sure to confirm the final results before continuing.

Then, filter the

output.yaml

(using the schema in

reference.md

) using the scripts provided to extract all the violations found.

IMPORTANT: If making a decision about the input directory, be sure to scope the input directory to only the relevant application code to avoid unnecessary processing of unrelated files.

Kantra Analyze Command

To show help for the Kantra analyze command, run:

kantra analyze --help

To run analysis on application source code, run:

kantra analyze --input=<path/to/source/code> --output=<path/to/output/report> --skip-static-report

If the output directory already exists, Kantra will fail. To overwrite the directory, supply the

--overwrite

flag.

[!IMPORTANT]

--skip-static-report

is used to avoid generating a static HTML report, which may not be needed in all contexts.

To list all sources and targets, use:

kantra analyze --list-sources
kantra analyze --list-targets

Add targets to the analyze command using multiple

--target=<target>

flags.

[!IMPORTANT] Try to limit the sources and targets to only those relevant to your application to avoid unnecessary processing of unrelated rules.

To list all providers, use:

kantra analyze --list-providers