ezl-keygraphandGitHub 86effd5240 feat(worker): record severity in analysis mode and fix prompt substitutions (#413) ...

* feat(worker): record severity in analysis mode alongside confidence

* fix(worker): align add_finding severity with the exploit collector's four levels

* refactor(worker): drop the dead REPORT_VULN_HEADING substitution

No prompt in the tree uses the placeholder, so the replacement was a no-op
on every render.

* fix(worker): strip all whitespace from TOTP secrets, not just the ends

* fix(worker): render rule type and value in the agent prompt

* refactor(worker): drop the dead vuln-summary subsection substitution

2026-08-04 21:02:00 +05:30

..

config-schema.json

feat(worker): record severity in analysis mode and fix prompt substitutions (#413)

2026-08-04 21:02:00 +05:30

example-config.yaml

feat: multi-provider model support, SARIF output, and exploit-mode fixes (#402)

2026-07-30 19:31:52 +05:30