# Chapter 4 — graph tour transcripts (captured)

Real output from the verification machine (same environment as the earlier transcripts). The Chapter 4 overlay (ch04-first-graph) adds cognitionzoo/critic/graph_tour_test.go — six tests covering build, seal, conditions, the visit cap, validation, and preflight — on top of the Chapter 2–3 package.

# Full critic package test run (ch02+ch03+ch04 overlays)

$ go test ./cognitionzoo/critic/... -count=1 -v
--- PASS: TestEnvelopeTour (0.00s)
--- PASS: TestFakeModelIsALanguageModel (0.00s)
--- PASS: TestGenerateStreamEmitsScriptedChunksInOrder (0.00s)
--- PASS: TestGenerateReturnsJoinedText (0.00s)
--- PASS: TestChatWithToolsReturnsScriptedToolCalls (0.00s)
--- PASS: TestTurnsCycleInOrder (0.00s)
--- PASS: TestGenerateStreamClosesOnCancellation (0.03s)
--- PASS: TestGenerateStreamDropsAndCountsOnSendContention (0.03s)
--- PASS: TestErrorInjectionFailsEveryMethod (0.00s)
--- PASS: TestEmptyScriptFailsCleanly (0.00s)
--- PASS: TestFirstGraphRuns (0.00s)
--- PASS: TestConditionalEdgesRoutes (0.00s)
--- PASS: TestGraphSealsOnExecute (0.00s)
--- PASS: TestVisitCapCatchesCycle (0.00s)
--- PASS: TestValidationCatchesBrokenWiring (0.00s)
--- PASS: TestPreflightGatesToolPlacement (0.00s)
PASS
ok  	codeburg.org/lexbit/relurpify/cognitionzoo/critic	0.059s

The six Chapter 4 tests, and what each asserts:

• TestFirstGraphRuns — intake (SystemNode) → drafter (hand-implemented Node, driven by FakeModel) → done (TerminalNode). Asserts the run’s result, the node-written envelope value (critic.draft), the runtime-captured result fields (drafter.draft), one model call, and an empty preflight report when no catalog is wired.
• TestConditionalEdgesRoutes — a ConditionFunc on drafter → polish routes on the draft’s content; the complementary edge to done is not taken.
• TestGraphSealsOnExecute — after Execute, AddNode and SetStart return ErrGraphSealed (asserted with errors.Is), and a second Execute still succeeds: sealing governs mutation, not execution.
• TestVisitCapCatchesCycle — a → b → a with SetMaxNodeVisits(2): Execute fails with potential cycle detected at node a.
• TestValidationCatchesBrokenWiring — SetStart("missing") → "start node missing not found"; AddEdge("done", "ghost", …) → "node ghost not defined"; Execute without a start → "graph has no start node".
• TestPreflightGatesToolPlacement — a ToolNode whose tool descriptor is tool:read_stringkit, against three states of the real capability/registry.CapabilityRegistry (the framework’s own unit-test pattern: NewRegistry + RegisterLegacyTool): not wired as catalog (report has no opinions), empty (capability_missing, blocking — Execute refuses), and with the tool registered (one placement, SelectedCapabilityID tool:read_stringkit, with its reason; Execute runs and the test asserts the invocation landed in the tool through the governed pipeline — tool.calls == 1). Invocations go through a governedInvoker adapter that forwards to the registry; there is no fake invoker in the book’s code.

# The verifier

$ ./verify.sh build ch04
== tier 1 (build): ch04 ==
ok  	codeburg.org/lexbit/relurpify/cognitionzoo/critic	0.059s
[PASS] build: ch04 compiles clean and its tests pass
$ ./verify.sh gates ch04
== tier 2 (gates): ch04 ==
... (build, vet, tests, make lint-arch — all PASS)
[PASS] gates: ch04 holds the framework's law