Code Health
Share your code and Seba assesses its quality — type safety, lint cleanliness, test coverage, dead code — then returns a health report with a weighted 0–10 score and fixes ranked by impact. It reports; it doesn’t change your code. Billed per run.
One honest score, with the fixes that move it
A weighted composite score
Type safety, lint, tests, dead code and script hygiene roll into one 0–10 number, weighted by what matters — so you see overall health, not five disconnected checks.
Fixes ranked by impact
Seba orders recommendations by weight times the gap to a 10, so you start with the change that lifts the score most, not the one that’s easiest to spot.
Honest about reality
A codebase with 100 type errors and all tests green is not healthy, and the score says so. No vanity grades — the number reflects what the checks actually found.
Just share the code
Paste files or send a public repo URL in Telegram. The report comes back as a file with the score and fix list, billed per run.
What the health check assesses
The same dimensions a staff engineer owning the CI dashboard tracks.
- Type safety — type errors and the holes that let bad data through
- Lint cleanliness — warnings, anti-patterns and style drift
- Test coverage — what’s tested, what only the happy path touches
- Dead code — unused exports, files and dependencies to remove
- Script hygiene — shell and build script issues
- Composite score — a weighted 0–10 across every dimension
- Per-dimension breakdown — score, status and the specific findings
- Ranked recommendations — the fixes that move the score most
From code to a health score in three steps
Open Seba in Telegram
One tap into @meetseba_bot. No install, no signup forms.
Share the code to assess
Paste files or send a public repo URL. “Assess the health of this module and score it.”
Get the report
Seba scores each dimension, computes the composite and returns a ranked fix list with the exact credit cost shown up front.
When to run a health check
Know where the codebase stands before it slips further.
- +Inheriting a codebase and sizing up its state
- +Before a refactor, to set a baseline score
- +Reviewing a module that keeps breaking
- +Deciding what tech debt to pay down first
- +Vetting an open-source dependency before adopting it
- +A quarterly quality check across the project
Frequently asked questions
What does the health check assess?+
Type safety, lint cleanliness, test coverage, dead code and script hygiene — rolled into a weighted 0–10 composite, with a per-dimension breakdown and ranked fixes.
Does Seba fix the issues for me?+
No. Seba returns a report with a score and prioritized fixes. You or your team apply them. Seba assesses what you share and does not change your code or push commits.
What does Seba assess from?+
Pasted code or a public repo URL. The more you share, the more complete the picture — Seba scores what it can see, and notes anything it couldn’t reach.
How is the score calculated?+
Each dimension is scored 0–10 and weighted — tests and types count most — into one composite. Dimensions it can’t assess are dropped and the weight redistributed, not penalized.
What format is the report?+
A structured report file with the composite score, a per-dimension table and a ranked fix list. Ask for a checklist or a summary and Seba formats it that way.
How much does it cost?+
You spend credits per run, and Seba shows the exact cost before it starts. Start free with welcome credits; credits never expire.
Score your code in a single message
Share your code with Seba and get a health report with a score and ranked fixes back. Pay only for the run.
No install · no signup forms