ci: ensure all declared AI TAP tests are built and executed

[renecannao]

Context

Both AI group definitions declare 22 tests, but the #6083 runs execute only 16 tests in ai-g1 and 17 tests in ai-g2.

The ubuntu24-tap-genai-gcov build sets SKIP_GENAI_UNIT_TESTS=1 and prunes GenAI unit binaries for disk usage, while the groups still register those binaries. The absent tests are therefore silently skipped rather than reported as skipped or assigned elsewhere.

Missing from ai-g1

• genai_config_query_unit-t
• genai_discovery_schema_unit-t
• genai_fts_string_unit-t
• genai_llm_clients_unit-t
• genai_mcp_endpoint_unit-t
• genai_mcp_thread_unit-t

Missing from ai-g2

• genai_mysql_catalog_unit-t
• genai_query_handler_unit-t
• genai_rag_fetch_from_source_unit-t
• genai_stats_parsing_unit-t
• genai_thread_unit-t

Acceptance criteria

• CI reports a declared-but-missing test binary as an error or explicit skip; it is never silently omitted.
• Every listed test is built and executed in a documented job, or removed from the AI integration group with a documented replacement.
• Group summaries reconcile declared, selected, skipped, and executed counts.
• Disk-usage constraints are handled without silently reducing coverage.
• The existing CI-unit-tests-asan-coverage workflow remains unchanged unless separately approved.

Parent: #6098
Observed in: #6083
Related layout discussion: #5735

[coderabbitai]

🔗 Related PRs

#5601 - ci: fix TAP build target so test binaries are actually produced (critical) [merged]
#5752 - Add config query, protobuf vendoring, and RAG source fetch [merged]
#5971 - ci(coverage): define Codecov flags by test binary class [merged]
#6024 - fix(build): make tap unit_tests wait for the tap library directory (intermittent 'libcurl.so: file too short') [merged]
#6085 - test: add targeted GCOV TAP workloads [merged]
#6089 - ci: wire AI TAP coverage shards [merged]
#6097 - fix(ci): hand off GenAI plugin to AI shards [open]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!