CHURNTwo-snapshot measurement
Quantifies added, deleted, changed, and unchanged lines between two versions of a project — per file and in aggregate.
In your pipeline
Most teams never open the app — it runs in CI, or on a schedule.
CodeDelta is a command-line engine first. Drop the GitHub Action into a repo and every pull request gets churn and AI findings posted back as a comment — with an optional gate that blocks the merge. Or run the same binary from cron for a nightly trend. Nothing is installed on a developer's machine, and the code never leaves your runner.
.github/workflows/codedelta.yml
name: CodeDelta
on: pull_request
jobs:
codedelta:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
with: { fetch-depth: 0 }
- uses: code-delta-app/action@v1
with:
mode: churn_agent # the default: churn + Agent Scan (both adds the ML audit)
fail-on-new: "true" # block merge on new findings
gate: "true" # block egress to CN/RU/KP/IR
bom: ai-bom.json # write an AI Bill of Materials
cron / batch — same engine, no GitHub
# nightly: diff the last commit, append to a trend DB
$ codedelta --git HEAD~1..HEAD --db trend.db --csv churn.csv
What it posts back on the pull request
The comment on every pull request. Churn (with REP_CHURN), AI audit and agent scan posted inline — plus the merge gate, which blocks on new findings when you turn it on.
…and the job that posts it. CodeDelta running inside the pipeline — fetch engine, run on the PR diff, post the comment — every step green, 18 seconds.
