test: altimate tools — impact analysis DAG traversal and training import parsing (#384)

* test: altimate tools — impact analysis DAG traversal and training import markdown parsing

New tests for two recently added tools (impact_analysis, training_import) that had
zero test coverage. These tools are user-facing and incorrect behavior leads to
wrong blast radius assessments or silent data loss when importing team standards.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

https://claude.ai/code/session_01H7d93hQP5qAwYhLgRdkqhV

* test: address critic review — boundary tests, tighter assertions, constant import

- Add severity boundary tests at exact thresholds (3→LOW, 4→MEDIUM, 10→MEDIUM, 11→HIGH)
- Fix blast radius test to verify model_count drives denominator, not models.length
- Import TRAINING_MAX_PATTERNS_PER_KIND instead of hardcoding 48
- Document that parseMarkdownSections includes empty sections (potential future fix)
- Tighten empty-section assertion from toBeGreaterThanOrEqual to exact toBe

https://claude.ai/code/session_01H7d93hQP5qAwYhLgRdkqhV

Author: anandgupta42

packages/opencode/test/altimate/impact-analysis.test.ts

@@ -0,0 +1,261 @@
+/**
+ * Tests for the impact_analysis tool — DAG traversal, severity classification,
+ * and report formatting.
+ *
+ * Mocks Dispatcher.call to supply known dbt manifests so we can verify
+ * findDownstream logic without a real napi binary or dbt project.
+ */
+import { describe, test, expect, spyOn, afterAll, beforeEach } from "bun:test"
+import * as Dispatcher from "../../src/altimate/native/dispatcher"
+import { ImpactAnalysisTool } from "../../src/altimate/tools/impact-analysis"
+import { SessionID, MessageID } from "../../src/session/schema"
+
+// Disable telemetry
+beforeEach(() => {
+  process.env.ALTIMATE_TELEMETRY_DISABLED = "true"
+})
+afterAll(() => {
+  delete process.env.ALTIMATE_TELEMETRY_DISABLED
+})
+
+const ctx = {
+  sessionID: SessionID.make("ses_test"),
+  messageID: MessageID.make("msg_test"),
+  callID: "call_test",
+  agent: "build",
+  abort: AbortSignal.any([]),
+  messages: [],
+  metadata: () => {},
+  ask: async () => {},
+}
+
+// Spy on Dispatcher.call so we control what "dbt.manifest" and "lineage.check" return
+let dispatcherSpy: ReturnType<typeof spyOn>
+
+function mockDispatcher(responses: Record<string, any>) {
+  dispatcherSpy?.mockRestore()
+  dispatcherSpy = spyOn(Dispatcher, "call").mockImplementation(async (method: string, params: any) => {
+    if (responses[method]) return responses[method]
+    throw new Error(`No mock for ${method}`)
+  })
+}
+
+afterAll(() => {
+  dispatcherSpy?.mockRestore()
+})
+
+describe("impact_analysis: empty / missing manifest", () => {
+  test("reports NO MANIFEST when manifest has no models", async () => {
+    mockDispatcher({
+      "dbt.manifest": { models: [], model_count: 0, test_count: 0 },
+    })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "stg_orders", change_type: "remove", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.title).toContain("NO MANIFEST")
+    expect(result.metadata.success).toBe(false)
+    expect(result.output).toContain("dbt compile")
+  })
+
+  test("reports MODEL NOT FOUND when model is absent", async () => {
+    mockDispatcher({
+      "dbt.manifest": {
+        models: [{ name: "dim_customers", depends_on: [], materialized: "table" }],
+        model_count: 1,
+        test_count: 0,
+      },
+    })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "stg_orders", change_type: "remove", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.title).toContain("MODEL NOT FOUND")
+    expect(result.metadata.success).toBe(false)
+    expect(result.output).toContain("dim_customers")
+  })
+})
+
+describe("impact_analysis: DAG traversal", () => {
+  const linearDAG = {
+    models: [
+      { name: "stg_orders", depends_on: [], materialized: "view" },
+      { name: "int_orders", depends_on: ["stg_orders"], materialized: "ephemeral" },
+      { name: "fct_orders", depends_on: ["int_orders"], materialized: "table" },
+      { name: "rpt_daily", depends_on: ["fct_orders"], materialized: "table" },
+    ],
+    model_count: 4,
+    test_count: 5,
+  }
+
+  test("finds direct and transitive dependents in a linear chain", async () => {
+    mockDispatcher({ "dbt.manifest": linearDAG })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "stg_orders", change_type: "remove", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.success).toBe(true)
+    expect(result.metadata.direct_count).toBe(1) // int_orders
+    expect(result.metadata.transitive_count).toBe(2) // fct_orders, rpt_daily
+    expect(result.output).toContain("int_orders")
+    expect(result.output).toContain("fct_orders")
+    expect(result.output).toContain("rpt_daily")
+    expect(result.output).toContain("BREAKING")
+  })
+
+  test("SAFE severity when no downstream models exist", async () => {
+    mockDispatcher({ "dbt.manifest": linearDAG })
+    const tool = await ImpactAnalysisTool.init()
+    // rpt_daily is a leaf — nothing depends on it
+    const result = await tool.execute(
+      { model: "rpt_daily", change_type: "modify", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.success).toBe(true)
+    expect(result.metadata.severity).toBe("SAFE")
+    expect(result.metadata.direct_count).toBe(0)
+    expect(result.output).toContain("safe to make")
+  })
+
+  test("diamond dependency counts each model only once", async () => {
+    mockDispatcher({
+      "dbt.manifest": {
+        models: [
+          { name: "src", depends_on: [], materialized: "view" },
+          { name: "left", depends_on: ["src"], materialized: "table" },
+          { name: "right", depends_on: ["src"], materialized: "table" },
+          { name: "merge", depends_on: ["left", "right"], materialized: "table" },
+        ],
+        model_count: 4,
+        test_count: 0,
+      },
+    })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "src", change_type: "rename", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.success).toBe(true)
+    // left, right are direct; merge is transitive. merge must appear only once.
+    expect(result.metadata.direct_count).toBe(2)
+    expect(result.metadata.transitive_count).toBe(1)
+    // Total = 3, which is LOW severity
+    expect(result.metadata.severity).toBe("LOW")
+  })
+
+  test("handles dotted depends_on references (e.g. project.model)", async () => {
+    mockDispatcher({
+      "dbt.manifest": {
+        models: [
+          { name: "stg_users", depends_on: [], materialized: "view" },
+          { name: "dim_users", depends_on: ["myproject.stg_users"], materialized: "table" },
+        ],
+        model_count: 2,
+        test_count: 0,
+      },
+    })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "stg_users", change_type: "retype", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.success).toBe(true)
+    expect(result.metadata.direct_count).toBe(1)
+    expect(result.output).toContain("dim_users")
+    expect(result.output).toContain("CAUTION") // retype warning
+  })
+})
+
+describe("impact_analysis: severity classification", () => {
+  function makeManifest(downstreamCount: number) {
+    const models = [{ name: "root", depends_on: [] as string[], materialized: "view" }]
+    for (let i = 0; i < downstreamCount; i++) {
+      models.push({ name: `model_${i}`, depends_on: ["root"], materialized: "table" })
+    }
+    return { models, model_count: models.length, test_count: 0 }
+  }
+
+  test("LOW severity boundary: exactly 3 downstream models", async () => {
+    mockDispatcher({ "dbt.manifest": makeManifest(3) })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "root", change_type: "modify", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.severity).toBe("LOW")
+  })
+
+  test("MEDIUM severity boundary: exactly 4 downstream models", async () => {
+    mockDispatcher({ "dbt.manifest": makeManifest(4) })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "root", change_type: "modify", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.severity).toBe("MEDIUM")
+  })
+
+  test("MEDIUM severity boundary: exactly 10 downstream models", async () => {
+    mockDispatcher({ "dbt.manifest": makeManifest(10) })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "root", change_type: "modify", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.severity).toBe("MEDIUM")
+  })
+
+  test("HIGH severity boundary: exactly 11 downstream models", async () => {
+    mockDispatcher({ "dbt.manifest": makeManifest(11) })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "root", change_type: "modify", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.metadata.severity).toBe("HIGH")
+  })
+})
+
+describe("impact_analysis: error handling", () => {
+  test("returns ERROR when Dispatcher throws", async () => {
+    mockDispatcher({}) // no mock for dbt.manifest — will throw
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "x", change_type: "remove", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    expect(result.title).toContain("ERROR")
+    expect(result.metadata.success).toBe(false)
+    expect(result.output).toContain("dbt compile")
+  })
+})
+
+describe("impact_analysis: blast radius percentage", () => {
+  test("percentage uses model_count, not models array length", async () => {
+    // model_count (20) intentionally differs from models.length (4)
+    // to verify the denominator comes from the declared count
+    mockDispatcher({
+      "dbt.manifest": {
+        models: [
+          { name: "root", depends_on: [], materialized: "view" },
+          { name: "child1", depends_on: ["root"], materialized: "table" },
+          { name: "child2", depends_on: ["root"], materialized: "table" },
+          { name: "unrelated", depends_on: [], materialized: "view" },
+        ],
+        model_count: 20,
+        test_count: 3,
+      },
+    })
+    const tool = await ImpactAnalysisTool.init()
+    const result = await tool.execute(
+      { model: "root", change_type: "remove", manifest_path: "target/manifest.json", dialect: "snowflake" },
+      ctx,
+    )
+    // 2 downstream out of 20 declared = 10.0%
+    expect(result.output).toContain("10.0%")
+    expect(result.output).toContain("2/20")
+  })
+})