Engineering intelligence · scryable.ai

Your codebase,
finally legible.

Pull from your git history. Surface who shipped what, how your team's velocity is really trending, and whether your AI tools are making the code better or worse.

Private beta · No card required · Works with local repos

Scroll
scryable · weekly report · api-service · 2026-04-01→07
Commits
284
↑ 12% vs last week
Contributors
9
Active this period
AI-assisted
22%
Cursor · Copilot
Churn rate
8%
↓ vs AI avg 14%
Commit velocity · 30 days
Authorship split
Human78%
AI-assisted22%

What scryable surfaces

Signal, not noise.
Data, not dashboards.

01
Per-developer output
Commits, line changes, velocity trends and commit size distributions — for every contributor, over any time window. No GitHub API required.
02
AI effectiveness tracking
Detect AI-assisted commits across Copilot, Cursor and Claude Code. Compare churn rates, revert frequency and code quality between AI and human-authored code.
Unique to scryable
03
Multi-repo aggregation
Point at any number of local or remote repos. Get a unified view across your entire engineering stack — not just one service in isolation.
04
Privacy-first, local by default
Runs entirely from your terminal against local clones. Your code never leaves your machine unless you explicitly choose to use AI summarisation.
05
AI commit summarisation
Each commit gets a concise, plain-English summary generated from the diff. Pluggable — use OpenAI, Anthropic or run offline with no AI at all.
06
Schedulable reports
Designed for cron and CI. Drop it into a GitHub Actions workflow and get a weekly Markdown report in your inbox or Slack — no dashboard required.

How it works

Three commands.
One clear picture.

Point at your repos
Pass local paths or remote Git URLs. Scryable clones, fetches, and resolves author identities automatically.
./scryable report \
  --repos ./api,./web \
  --since 2026-03-01
Scryable reads the history
Git log is parsed, authors resolved, diffs annotated, and AI-assisted commits detected — all locally, in seconds.
Collecting commits...
Resolved 9 authors
Detected 22% AI-assisted
Get your report
Markdown, JSON, or open the dashboard. Share with your team, feed into CI, or drop into your weekly standup.
Wrote reports/
  api-20260301-07.md
  api-20260301-07.json

The AI tracking layer

Is your AI
actually helping?

63% of developers use AI coding tools. But trust in AI-generated code dropped to 29% in 2025. Engineering leaders are flying blind — spending on Copilot and Cursor without any data on whether it's making their codebase better or worse.

Scryable detects AI-assisted commits across every major tool, then compares churn rate, revert frequency, and commit size distributions between AI and human-authored code. For the first time, you can answer the question with data.

29%
Trust in AI-generated code, 2025
4×
Increase in duplicate code blocks from AI
63%
Developers using AI tools weekly
0
Teams with real visibility into the impact

Private beta

See what your team
is really building.

Join the early access list. First 200 teams get lifetime free tier.

Works with GitHub · GitLab · Bitbucket · any local repo