datadog/techdebt
>- Review a code diff / branch / PR for technical debt — code duplication, unnecessary complexity / over-engineering, and redundant or dead code. Use whenever the user wants a tech-debt, cleanup, or refactor review, asks to check a branch or PR for duplication / complexity / dead code before opening a PR, or behavior-preserving fixes; it never changes behavior.
npx skills add https://github.com/DataDog/dd-trace-java --skill techdebt
Analyze changes on the current branch to identify and fix technical debt, code duplication, and unnecessary complexity.
Find the merge-base (where this branch diverged from master) and compare against it:
# Find upstream (DataDog org repo)
UPSTREAM=$(git remote -v | grep -E 'DataDog/[^/]+(.git)?\s' | head -1 | awk '{print $1}')
if [ -z "$UPSTREAM" ]; then
echo "No DataDog upstream found, using origin"
UPSTREAM="origin"
fi
# Find the merge-base (commit where this branch diverged from master)
MERGE_BASE=$(git merge-base HEAD ${UPSTREAM}/master)
echo "Comparing changes introduced on this branch since diverging from master using base commit: $MERGE_BASE"
git diff $MERGE_BASE --stat
git diff $MERGE_BASE --name-status
If no changes exist, inform the user and stop.
If changes exist, read the diff and the full content of modified source files (not test files) to understand context.
Look for:
Code Duplication
Unnecessary Complexity
Redundant Code
Present a concise summary of issues found with file:line references.
Then ask the user if they want you to fix the issues. When fixing:
Take datadog/techdebt from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.