Know what's
actually built.
Waterline reads your codebase and scores every acceptance criterion against merged code. What shipped, what is missing, any ticket, any point in time.
Acceptance Criteria
Acceptance Criteria
Agents write code at 10x.
Traceability just took a massive hit.
The bottleneck is no longer writing code. It is knowing what actually got built. When agents verify their own work, reliability decreases. Waterline reads the code independently.
Requirements, not just commits
Waterline evaluates whether each acceptance criterion was satisfied in the code and where.
Requirement coverage at sprint scale
Before the sprint review, you see which requirements landed in code and which did not.
Agent commits read the same as human commits
Waterline reads evidence in the code. Who wrote the commit does not change what it satisfies.
acme/payments
Recent commits
feat: implement exponential backoff for payment retry
fix: dead letter queue on retry exhaustion
chore: add retry config to env template
Sprint intelligence
from the code.
Waterline reads every merged commit and scores each acceptance criterion. You see what landed before the review.
Criterion-level analysis
Each acceptance criterion evaluated against merged commits. See exactly which requirements are done, which are partial, and which have no coverage.
SRV-114 · acme/auth-service
Token refresh flow
4 commits analyzed · b72dc1a
Sprint health before the review
See which tickets are on track, at risk, or stalled before anyone has to ask.
Sprint 31
Sprint Health
Spec quality check
Waterline flags vague acceptance criteria before analysis runs and suggests a concrete rewrite for each. You fix the spec before the code goes wrong.
BIL-88 · Spec check
Acceptance Criteria Quality
Users should be able to export their data
Suggested: Export all user records as UTF-8 CSV within 30 seconds for datasets up to 50k rows.
CSV export includes name, email, created_at, plan
Handle large exports without timing out
Suggested: Queue exports > 10k rows async. Deliver via email link within 5 minutes.
Running in under
five minutes.
No infrastructure to manage. No agents to configure.
Connect GitHub and Jira
No code to deploy and no configuration files to write.
Works with Jira Cloud, GitHub Issues, GitHub, and Slack.
Waterline indexes your codebase
Select a repo and branch. Waterline indexes every commit and pull request against your codebase.
Initial index takes a few minutes for large repos. Every new commit syncs automatically after that.
Analyze any ticket in seconds
Open a ticket and click analyze. Waterline matches the acceptance criteria to your merged code and returns a completion score.
Each result shows the commits it's based on. Re-analyze any time.
MAC-247 · Spec quality check
Acceptance Criteria Analysis
The payment system should handle errors gracefully
Issue: No measurable outcome defined
Suggested: Return HTTP 503 with Retry-After header on gateway timeout. Retry up to 3 times before failing to dead-letter queue.
Implement exponential backoff with jitter on retry attempts
Add proper logging to the retry flow
Issue: Scope and format not specified
Suggested: Log each retry attempt to structured JSON with fields: attempt_number, delay_ms, error_code, request_id.
Bad acceptance criteria cost more
than bad code.
Vague specs produce wrong implementations. Before any code is written, Waterline tells you which criteria are testable and rewrites the ones that are not.
Testable vs. vague, flagged automatically
Waterline classifies every acceptance criterion before analysis. Testable criteria produce reliable results. Vague ones get flagged with an explanation of what's missing.
Suggested rewrites, not just warnings
For each vague criterion, Waterline generates a concrete rewrite that leads to accurate, testable criteria.
Write better specs over time
When you see which criteria Waterline can evaluate reliably, you write tighter specs for the next sprint.
Request early access.
Waterline is in private early access. Request your spot to lock in launch pricing and help shape what it will become before it goes public.
What's included
- GitHub + Jira Cloud integration
- GitHub Issues support
- Symbol-level code indexing
- Per-criterion analysis
- Acceptance criteria quality check
- Sprint health overview
- Slack notifications
- Launch pricing locked in
FAQ
Something else? Email us.