fix: wire allowed_tools/disallowed_tools inputs through to Claude SDK

The disallowed tools env var read in run.ts was DISALLOWED_TOOLS_INPUT
(raw user input) instead of DISALLOWED_TOOLS (computed value including
WebSearch/WebFetch defaults). This meant default tool disallowance was
never enforced. Also removes orphaned ALLOWED_TOOLS env var export from
createPrompt() (tag mode passes allowed tools via claudeArgs instead),
extracts shared parseToolsInput utility to eliminate duplication, and
adds tests.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

action.yml

@@ -77,6 +77,14 @@ inputs:
     required: false
     default: "false"
 
+  allowed_tools:
+    description: "Comma-separated list of additional tools to allow (e.g., 'WebSearch,WebFetch,Bash(npm test:*)'). These are merged with the default tag-mode tools."
+    required: false
+    default: ""
+  disallowed_tools:
+    description: "Comma-separated list of tools to disallow (e.g., 'WebSearch'). Overrides defaults."
+    required: false
+    default: ""
   claude_args:
     description: "Additional arguments to pass directly to Claude CLI"
     required: false
@@ -254,6 +262,9 @@ runs:
         INCLUDE_FIX_LINKS: ${{ inputs.include_fix_links }}
         ADDITIONAL_PERMISSIONS: ${{ inputs.additional_permissions }}
         CLAUDE_ARGS: ${{ inputs.claude_args }}
+        # _INPUT suffix = raw user input; createPrompt() computes DISALLOWED_TOOLS from these
+        ALLOWED_TOOLS_INPUT: ${{ inputs.allowed_tools }}
+        DISALLOWED_TOOLS_INPUT: ${{ inputs.disallowed_tools }}
         ALL_INPUTS: ${{ toJson(inputs) }}
 
         # Base-action inputs