feat(ai): prompt-injection delimiters + synonym-aware match scoring

Wave 4 hardening. Wrap untrusted CV/job-description/instruction text
in tools/summarizer prompts with explicit delimiters and an
ignore-embedded-instructions rule, since JD text, recruiter emails,
and free-text candidate background all flow into rewrite/normalize
prompts unescaped today.

Match score previously normalized synonyms (JS/Kubernetes/K8s/etc)
only when scanning the job posting, not when checking the CV corpus,
so a CV using an abbreviation the job spelled out never matched.
SkillTagger.MatchesTag reuses the same synonym regex for both sides.

JobTrackerApi.Tests/JobCvMatchServiceTests.cs

@@ -90,6 +90,19 @@ public sealed class JobCvMatchServiceTests
         Assert.Equal(0, result.MatchedCount);
     }
 
+    [Fact]
+    public void Curated_tag_matches_synonym_spelling_in_cv()
+    {
+        // Job posting says "Kubernetes"; CV only says "K8s" -- same skill, different spelling.
+        var result = _service.Evaluate(
+            jobTitle: "Platform Engineer",
+            jobText: "Deep Kubernetes experience required for our platform team.",
+            cvSections: Sections(("Skills", "K8s, Terraform, Helm")));
+
+        Assert.Contains("Kubernetes", result.MatchedKeywords);
+        Assert.DoesNotContain("Kubernetes", result.MissingKeywords);
+    }
+
     [Fact]
     public void Title_keywords_are_weighted_and_missing_ones_rank_first()
     {