refactor: split dependency changes out of the security PR

The codex optional extra and the vllm/datasets floor bumps are dependency
hygiene / CVE-floor changes, not part of the command-injection and path-
injection hardening. Keep this PR surgical (the four security fixes + WebUI
tests); the dependency changes are preserved in the branch history (commit
6f0030d) for a separate dependency PR. The gradio floor comment stays as-is
(already synced to pyproject's 5.50.0 floor).

Author: WODE25500

pyproject.toml

@@ -38,12 +38,10 @@ dependencies = [
 alfworld = ["alfworld>=0.4.0", "gymnasium>=0.29.0"]
 # Claude model backend
 claude = ["claude-agent-sdk>=0.1.0", "json_repair>=0.61.0"]
-# Codex model backend (via OpenAI Codex SDK)
-codex = ["openai-codex-sdk>=0.1.0"]
 # Qwen local model backend (via vLLM)
-qwen = ["vllm>=0.8.4", "json_repair>=0.61.0"]
+qwen = ["vllm>=0.4.0", "json_repair>=0.61.0"]
 # SearchQA data materialization
-searchqa = ["datasets>=3.0"]
+searchqa = ["datasets>=2.18.0"]
 # Documentation site
 docs = ["mkdocs-material>=9.5.0", "mkdocstrings[python]>=0.24.0"]
 # WebUI dashboard

requirements.txt

@@ -15,7 +15,7 @@ httpx>=0.27.0
 # claude-agent-sdk>=0.1.0
 
 # ── Optional: Qwen local model (via vLLM) ────────
-# vllm>=0.8.4
+# vllm>=0.4.0
 
 # ── Optional: tolerant JSON repair for free-form output from non-OpenAI
 #    backends (Claude/Qwen). Without it extract_json() falls back safely and