feat(research): wire tools + research workflow graph (research-workflow §2/§3)

Makes the research feature runnable end-to-end, off by default. - config: [tools.research] (enabled, searxng_url, max_results, max_fetch_bytes). - registration: web_search/web_fetch are built into BOTH the default and per-workspace tool registries when research.enabled, sharing one HTTP client threaded from Main (none built on the static path). Egress stays harness-enforced: web_fetch is T2 (operator-approved) and the existing NetworkHostRule still applies. - workflow: examples/workflows/research.toml — decompose → gather → report, with the three artifact schemas and prompts. Fan-out (search per sub-question, fetch per source) runs as repeated tool calls inside the gather stage (Correx has no parallel agents); per-source synthesis into the dossier is the compression step, so the report stage consumes summaries, never raw pages. ResearchWorkflowTest validates the graph contract. To run: set [tools.research].enabled, register the 3 [[artifacts]], copy research.toml + prompts + schemas into the workflows dir, start SearXNG. Launch like any workflow (the T2 fetch approval surfaces as an approval card; the report opens in the artifact viewer). Follow-ups (noted, not blocking): batch fetch-approval at the source-list level (§3), a dedicated SourceFetched/LowQualityExtraction event (quality + content hash are already in tool-result metadata), dynamic per-session egress allowlist, and the web approval client (§6).
2026-06-13 23:18:19 +04:00
parent ad2d38ce46
commit 7a0d4d0ee2
11 changed files with 321 additions and 1 deletions
@@ -0,0 +1,53 @@
+package com.correx.infrastructure.workflow
+
+import com.correx.core.artifacts.kind.ConfigArtifactKind
+import com.correx.core.artifacts.kind.DefaultArtifactKindRegistry
+import com.correx.core.artifacts.kind.JsonSchema
+import org.junit.jupiter.api.Assertions.assertEquals
+import org.junit.jupiter.api.Assertions.assertTrue
+import org.junit.jupiter.api.Test
+import java.nio.file.Path
+import kotlin.io.path.exists
+
+/**
+ * Validates that the shipped research workflow (examples/workflows/research.toml) parses into a
+ * structurally valid WorkflowGraph: reachable stages, every `needs` produced upstream, the gather
+ * stage wired to the two network tools. Synthesis quality is inference-dependent and out of scope;
+ * this guards the graph contract.
+ */
+class ResearchWorkflowTest {
+
+    private val registry = DefaultArtifactKindRegistry().apply {
+        listOf("research_plan", "source_dossier", "research_report").forEach { id ->
+            register(ConfigArtifactKind(id = id, schema = JsonSchema(type = "object"), llmEmitted = true))
+        }
+    }
+
+    private fun repoFile(relative: String): Path {
+        var dir: Path? = Path.of("").toAbsolutePath()
+        while (dir != null) {
+            val candidate = dir.resolve(relative)
+            if (candidate.exists()) return candidate
+            dir = dir.parent
+        }
+        error("could not locate $relative from ${Path.of("").toAbsolutePath()}")
+    }
+
+    @Test
+    fun `research workflow builds a valid graph`() {
+        val graph = TomlWorkflowLoader(registry).load(repoFile("examples/workflows/research.toml"))
+
+        assertEquals("research", graph.id)
+        assertEquals("decompose", graph.start.value)
+        assertEquals(setOf("decompose", "gather", "report"), graph.stages.keys.map { it.value }.toSet())
+
+        val gather = graph.stages.values.first { it.produces.any { a -> a.name.value == "source_dossier" } }
+        assertEquals(setOf("research_plan"), gather.needs.map { it.value }.toSet())
+        assertEquals(setOf("web_search", "web_fetch"), gather.allowedTools)
+
+        val report = graph.stages.values.first { it.produces.any { a -> a.name.value == "research_report" } }
+        assertEquals(setOf("source_dossier"), report.needs.map { it.value }.toSet())
+
+        assertTrue(graph.transitions.any { it.to.value == "done" }, "must terminate at done")
+    }
+}