| name | style-check |
| description | Run deterministic style checks on a documentation file or directory, then layer AI judgment on top using the migrated style-checker prompt. |
| argument-hint | [--fix] [--no-log] <file or directory path> |
| user-invocable | true |
/inki:style-check: style lint for documentation files
Step 1: Determine target
$ARGUMENTS is a relative path. If empty, use the diff between current branch and main:
git diff main...HEAD --name-only -- '*.md'
Logging: unless --no-log is passed, write this skill's report to the run log per ../../references/logging.md (--log-dir <path> and --short-log are also accepted). When invoked as part of an orchestrator (e.g. /inki:review), write into that run's existing directory instead of creating a new one.
Step 2: Run the deterministic style script
The plugin ships a deterministic linter at claude-plugins/inki/scripts/style-lint.sh. It checks Markdown/MDX files against regex-based rules from the 12 Rules of Technical Writing (em dashes, casual language, simplification suggestions, bold prefixes that should be admonitions, etc.). Exit codes: 0 clean, 1 errors found, 2 warnings only.
Resolve the script path from the skill's location and run it on the target:
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]:-$0}")/../../scripts" && pwd)"
"$SCRIPT_DIR/style-lint.sh" <target>
If invoked from a Claude Code session, the equivalent is to read ../../scripts/style-lint.sh relative to this SKILL.md and execute it on the target.
Step 3: Apply the migrated style-checker prompt
Read ../../references/prompts/style-checker.md (relative to this SKILL.md location) and use it as the system prompt for an AI judgment pass over the target file's content.
Step 4: Report
Output findings in this format:
File: <path>
Deterministic issues:
- <line>: <issue>
AI-judged issues:
- <issue>: <suggested fix>
If invoked with --fix in $ARGUMENTS, apply non-controversial fixes (typos, formatting) and leave the rest as suggestions.
