fix(inference,tools): unblock llama.cpp tool stages + workspace-relative file_read

Two fixes surfaced by the 2026-06-08 QA audit running role_pipeline on a local
llama-server.

F-001: LlamaCppInferenceProvider sent both a GBNF grammar and tools on the same
request; llama.cpp rejects the combination ("Cannot use custom grammar
constraints with tools"), failing every stage that has tools and produces a
schema-constrained artifact. Drop the grammar when tools are present and rely on
post-hoc schema validation + retry (invariant #7 still holds). TODO left for a
first-class emit_artifact tool as the proper fix.

F-005: FileReadTool resolved relative paths via toAbsolutePath() against the JVM
process cwd instead of the bound workspace, so the file-read jail anchored to the
wrong root and disagreed with the Plane-2 PATH_CONTAINMENT check. Add workingDir
to FileReadConfig, wire workspace.workingDir in the per-workspace tool builder,
and give FileReadTool a resolvePath() mirroring FileWriteTool/FileEditTool.

infrastructure/tools/src/main/kotlin/com/correx/infrastructure/tools/ToolConfig.kt

@@ -14,7 +14,11 @@ data class ToolConfig(
     val fileEdit: FileEditConfig = FileEditConfig(),
 )
 
-data class FileReadConfig(val enabled: Boolean = false, val allowedPaths: Set<Path> = emptySet())
+data class FileReadConfig(
+    val enabled: Boolean = false,
+    val allowedPaths: Set<Path> = emptySet(),
+    val workingDir: Path? = null,
+)
 data class FileWriteConfig(
     val enabled: Boolean = false,
     val allowedPaths: Set<Path> = emptySet(),
@@ -49,6 +53,7 @@ fun ToolConfig.buildTools(): List<Tool> = buildList {
         add(
             FileReadTool(
                 allowedPaths = fileRead.allowedPaths,
+                workingDir = fileRead.workingDir,
             ),
         )
     }