Walker Baselines and Diffs
Baselines und diffs compare two autonomous walks for same application. Help separate new, persistent und resolved findings und show where recorded UI state changed.
Baseline ist comparison reference, not certification. Pin only after reviewing target, authentication state, limits, findings und coverage.
Recommended workflow
- Run und review known-good walk.
- Pin with
set_walk_baseline. - Deploy change und run new walk with comparable configuration.
- Call
diff_walkswith two walk IDs. - Inspect recursive report und every blocker/major finding.
- Run saved scenarios for flows where deterministic coverage matters.
set_walk_baseline(
walk_id="<reviewed-walk>",
app_name="my-app",
notes="Reviewed staging reference for revision <sha>"
)
diff_walks(
base_walk_id="<reviewed-walk>",
new_walk_id="<candidate-walk>"
)Identity und comparison limits
States content-addressed from route und DOM signatures. Findings joined with normalized deduplication keys. Makes stable comparisons possible, dynamic content can still make one logical screen look new, normalization can occasionally merge or split findings unexpectedly.
For defensible comparison keep inputs aligned:
- application und environment;
- authenticated account und role;
- seed URL und allowed origins;
- walker mode, state/depth/action limits und fuzz profile; und
- feature flags und test data.
If inputs differ, document difference instead of treating counts as release-quality regression signal.
Available tools
| Tool | Purpose |
|---|---|
set_walk_baseline | Pin one reviewed walk per owner/app key |
get_walk_baseline / list_walk_baselines | Inspect current references |
clear_walk_baseline | Remove pin without deleting walk |
diff_walks | Compare states und findings across two owned walks |
walk_recursive_report | Render recorded state tree with change annotations |
walk_findings_to_tickets | Produce tracker payloads for selected findings |
run_all_app_scenarios | Queue saved scenarios for app |
Ticket export returns JSON payloads for Linear, Jira oder GitHub. Posts them nicht. Review titles, evidence links, severity, sensitive content, project und assignee before sending to external tracker.
Release interpretation
Useful candidate signal:
- no new blocker or major finding in comparable coverage;
- targeted finding appears resolved;
- important saved scenarios pass in intended environment; und
- someone reviews any removed states or reduced coverage.
Call release green solely because findings_new zero nicht. Failed login, early timeout, smaller frontier oder
unavailable tester can also produce fewer findings.