slim the submit schemas: evidence and messages die; stringified-payload heal at the pre-validation seam

OpenRouter log diagnosis (owner, 2026-09-28): extraction time was dominated by tool-validation retries. The failing call had the whole propositions array emitted as ONE JSON-STRING ('propositions.0: must be object'), a truncated placeholder artifact inside it, and word-salad statements late in the array — payload-size collapse. Field audit: 'evidence' was never persisted anywhere and its faithfulness-checking consumer died with the fully-trusted ruling; 'messages' fed messageId granularity that NOTHING downstream reads, and the model sent [] anyway (chat-level provenance in practice). 'aliases' stays (optional now): the auditor's find tool globs them. Schema: propositions = {id, name, statement, aliases?}; edges = {from, to, type}. Prompts lose the evidence/messages doctrine and [id]-anchor guidance, gain a compact faithfulness line; transcriptView renders anchorless. The auditor echo instruction drops the evidence/messages mention, so its payload halves too. Heal: healStringifiedArrays on the submit tool's prepareArguments (pi-agent-core's sanctioned pre-TypeBox shim) parses a stringified propositions/edges back into arrays — the LaTeX-restore doctrine: heal weak-model wire behavior at the seam. Test drives the real agent-loop seam with a stringified submit.
2026-09-28 19:04:39 +03:00
parent 14a7f0de1d
commit bd9694b0e4
7 changed files with 131 additions and 125 deletions
@@ -221,7 +221,7 @@ after(() => {
 });

 /** What the extractor would produce for the fixture session — tree-blind,
- *  free-form, with message-granular provenance. */
+ *  free-form, with chat-level provenance. */
 const FINDINGS: SessionFindings = {
 	propositions: [
 		{
@@ -229,16 +229,12 @@ const FINDINGS: SessionFindings = {
 			name: "Derivative",
 			statement: "The derivative measures how fast a function changes",
 			aliases: [],
-			messages: ["a2"],
-			evidence: ["The derivative measures how fast a function changes."],
 		},
 		{
 			id: "p2",
 			name: "Chain rule",
 			statement: "The chain rule differentiates composed functions by multiplying local rates.",
 			aliases: [],
-			messages: ["a2"],
-			evidence: ["teach me the chain rule"],
 		},
 	],
 	edges: [
@@ -246,8 +242,6 @@ const FINDINGS: SessionFindings = {
 			from: "p1",
 			to: "p2",
 			type: "follows",
-			messages: ["a2"],
-			evidence: ["teach me the chain rule"],
 		},
 	],
 };
@@ -322,20 +316,21 @@ function silentModel(): PiEngine {
 // ─── transcriptView (pure) ───────────────────────────────────────────────────

 describe("transcriptView (pure — the shared markdown engine, owner 2026-09-27)", () => {
-	it("renders id-anchored speaker blocks + check blocks — tree-blind, plan ops folded", () => {
+	it("renders anchorless speaker blocks + check blocks — tree-blind, plan ops folded", () => {
 		const { record } = fixtureChat();
 		const view = transcriptView(record);

 		// The document header is the session title.
 		assert.match(view, /^# Extraction fixture$/m);

-		// Conversation verbatim, each block headed by its speaker + the
-		// message-id anchor — the provenance the extractor reports back.
-		assert.match(view, /### Learner · u1\n\nteach me the chain rule/);
-		assert.match(view, /### Teacher · a2\n\nThe derivative measures how fast a function changes\./);
+		// Conversation verbatim, each block headed by its speaker only —
+		// the [id] anchors are GONE (submit-schema slim, 2026-09-28:
+		// propositions no longer report message ids).
+		assert.match(view, /### Learner\n\nteach me the chain rule/);
+		assert.match(view, /### Teacher\n\nThe derivative measures how fast a function changes\./);

 		// Checks render as readable blocks (question + outcome).
-		assert.match(view, /### Quiz · r_probe-1/);
+		assert.match(view, /### Quiz\n/);
 		assert.match(view, /\*\*Q:\*\* What is a derivative, roughly\?/);
 		assert.match(view, /The learner answered incorrectly\./);
 		assert.match(view, /The learner answered correctly\./);
@@ -373,6 +368,25 @@ describe("transcriptView (pure — the shared markdown engine, owner 2026-09-27)
 // ─── the pipeline ───────────────────────────────────────────────────────────

 describe("the extraction pipeline (extractor → auditor → merger)", () => {
+	it("stringified-payload heal (owner, 2026-09-28): propositions arriving as ONE JSON string still validate and merge", async () => {
+		const { chatId } = fixtureChat();
+		const ledgers = memoryLedgers();
+		const verdict: AuditResult = { propositions: FINDINGS.propositions, edges: FINDINGS.edges };
+
+		// The weak-model failure from the OpenRouter log: the whole array
+		// emitted as a JSON-STRING ("propositions.0: must be object").
+		// pi-agent-core's prepareArguments shim runs BEFORE TypeBox — the
+		// heal parses it, validation passes, no retry burned.
+		const extractor = submitModel("submit_findings", {
+			propositions: JSON.stringify(FINDINGS.propositions),
+			edges: FINDINGS.edges,
+		});
+
+		const pipeline = createExtractionPipeline({ chatStore, graphStore: ledgers, extractor, auditor: submitModel("submit_review", verdict) });
+		const stats = await pipeline.runExtraction(chatId);
+		assert.deepEqual(stats, { nodes: 2, edges: 1 });
+	});
+
 	it("runs end to end: findings → audit verdict (identity) → deterministic merge + checkpoint", async () => {
 		const { chatId } = fixtureChat();
 		const ledgers = memoryLedgers();
@@ -631,7 +645,7 @@ describe("the extraction pipeline (extractor → auditor → merger)", () => {
 		const ledgers = memoryLedgers();

 		const single = (id: string, name: string, statement: string): SessionFindings => ({
-			propositions: [{ id, name, statement, aliases: [], messages: ["a2"], evidence: ["e"] }],
+			propositions: [{ id, name, statement, aliases: [], }],
 			edges: [],
 		});
 		const findingsA = single("pa", "Alpha claim", "Alpha statement, full precision");
@@ -688,7 +702,7 @@ describe("the extraction pipeline (extractor → auditor → merger)", () => {
 		const ledgers = memoryLedgers();

 		const single = (id: string, name: string, statement: string): SessionFindings => ({
-			propositions: [{ id, name, statement, aliases: [], messages: ["a2"], evidence: ["e"] }],
+			propositions: [{ id, name, statement, aliases: [], }],
 			edges: [],
 		});

@@ -938,8 +952,6 @@ const NOTES_FINDINGS = {
 			name: "Vim normal mode",
 			statement: "Vim starts in Normal mode, where keys are commands.",
 			aliases: [],
-			messages: [],
-			evidence: ["Vim always starts in Normal mode — keys are commands, not text."],
 		},
 	],
 	edges: [],
@@ -980,11 +992,12 @@ describe("the notes front door (importNotes — format-agnostic)", () => {
 		assert.equal(source.extractions![0]!.stats?.nodes, 1);
 		assert.ok(source.extractions![0]!.stages.extractor.steps.length > 0, "faux usage captured");

-		// File-scoped provenance: the mapping line carries the source id;
-		// the node carries NO message anchors.
+		// File-scoped provenance (submit-schema slim, 2026-09-28): the
+		// node carries ONE chat-level point — the source key itself, no
+		// message anchors.
 		const node = ledgers.readNodes()[0]!;
 		assert.equal(node.statement, "Vim starts in Normal mode, where keys are commands.");
-		assert.deepEqual(node.provenance, []);
+		assert.deepEqual(node.provenance, [{ chatId: source.id, messageId: "" }]);
 		assert.deepEqual(ledgers.readMappings(), [{ chatId: source.id, sourceId: "p1", graphNodeId: node.id }]);
 		// The checkpoint stamps the import.
 		assert.equal(ledgers.readCheckpoints()[0]!.chatId, source.id);
@@ -1102,10 +1115,10 @@ describe("the chunked pipeline (extraction coverage)", () => {
 		// Two batches: the first held messages 1-3, the second 4-5 — a
 		// message is never split, batches are consecutive oldest-first.
 		assert.equal(views.length, 2, "one extractor call per batch");
-		assert.ok(views[0]!.includes("### Learner · u1") && views[0]!.includes("### Learner · u3"));
-		assert.ok(!views[0]!.includes("### Learner · u4"), "the batch stops at the token budget");
-		assert.ok(views[1]!.includes("### Learner · u4") && views[1]!.includes("### Learner · u5"));
-		assert.ok(!views[1]!.includes("### Learner · u3"));
+		assert.ok(views[0]!.includes("### Learner") && views[0]!.includes("topic 1:"));
+		assert.ok(!views[0]!.includes("topic 4:"), "the batch stops at the token budget");
+		assert.ok(views[1]!.includes("### Learner") && views[1]!.includes("topic 4:"));
+		assert.ok(!views[1]!.includes("topic 3:"));
 		// The document header is the session title.
 		assert.ok(views[0]!.startsWith("# Chunky coverage fixture"));

@@ -1184,7 +1197,7 @@ describe("the chunked pipeline (extraction coverage)", () => {
 		});
 		await finisher.runExtraction(chatId);
 		assert.equal(reRunViews.length, 1);
-		assert.ok(reRunViews[0]!.includes("### Learner · u4") && reRunViews[0]!.includes("### Learner · u5"));
+		assert.ok(reRunViews[0]!.includes("### Learner") && reRunViews[0]!.includes("topic 4:"));
 		assert.ok(!reRunViews[0]!.includes("### Learner · u1"), "the covered remainder never re-enters the prompt");
 		assert.equal(ledgers.readCoverage().length, 5, "the re-run closed the gap");
 	});
@@ -224,24 +224,18 @@ function fixtureAudit(): AuditResult {
 				name: "Alpha",
 				statement: "Alpha intuition",
 				aliases: ["alpha feel"],
-				messages: ["a1", "a3"],
-				evidence: ["fixture evidence"],
 			},
 			{
 				id: "p2",
 				name: "Beta rule",
 				statement: "Beta rule",
 				aliases: [],
-				messages: ["a4"],
-				evidence: ["fixture evidence"],
 			},
 			{
 				id: "p3",
 				name: "Gamma",
 				statement: "Gamma application",
 				aliases: [],
-				messages: ["a5"],
-				evidence: ["fixture evidence"],
 			},
 		],
 		edges: [
@@ -249,8 +243,6 @@ function fixtureAudit(): AuditResult {
 				from: "p1",
 				to: "p2",
 				type: "follows",
-				messages: ["a4"],
-				evidence: ["fixture evidence"],
 			},
 		],
 	};
@@ -282,7 +274,7 @@ function scopedGraph(chatIds: string[]) {
 }

 describe("the merger (mergeFindings)", () => {
-	it("creates nodes from the verdicts with message-granular provenance", () => {
+	it("creates nodes from the verdicts with chat-level provenance", () => {
 		const stats = mergeFindings(chat1.id, fixtureAudit(), graphDeps, sessionSource(chat1.id));
 		assert.deepEqual(stats, { nodes: 3, edges: 1 });

@@ -293,14 +285,13 @@ describe("the merger (mergeFindings)", () => {
 			assert.match(node.id, /^n_/);
 			assert.ok(node.name.length > 0, "name is required");
 			assert.equal("kind" in node, false, "no kind field");
-			assert.ok(Array.isArray(node.provenance), "message-granular provenance");
+			assert.ok(Array.isArray(node.provenance), "provenance attached");
 		}
-		// Provenance is message-granular: p1 came up in a1 and a3.
+		// Provenance is CHAT-LEVEL (submit-schema slim, 2026-09-28):
+		// propositions no longer carry message ids — the run's source key
+		// is the whole provenance.
 		const alpha = graph.nodes.find((n) => n.statement === "Alpha intuition")!;
-		assert.deepEqual(alpha.provenance, [
-			{ chatId: chat1.id, messageId: "a1" },
-			{ chatId: chat1.id, messageId: "a3" },
-		]);
+		assert.deepEqual(alpha.provenance, [{ chatId: chat1.id, messageId: "" }]);

 		// The follows edge, provenance attached.
 		assert.equal(graph.edges.length, 1);
@@ -308,7 +299,7 @@ describe("the merger (mergeFindings)", () => {
 		assert.equal(edge.from, idByStatement.get("Alpha intuition"));
 		assert.equal(edge.to, idByStatement.get("Beta rule"));
 		assert.equal(edge.type, "follows");
-		assert.deepEqual(edge.provenance, [{ chatId: chat1.id, messageId: "a4" }]);
+		assert.deepEqual(edge.provenance, [{ chatId: chat1.id, messageId: "" }]);

 		// Mapping lines: one per (chat, sourceId).
 		assert.deepEqual(
@@ -351,8 +342,6 @@ describe("the merger (mergeFindings)", () => {
 						name: "Alpha",
 						statement: "Alpha feel", // a DIFFERENT phrasing — identity is the verdict, not strings
 						aliases: ["alpha intuition"],
-						messages: ["a1"],
-						evidence: ["fixture evidence"],
 						linkTo: alphaId,
 					},
 				],
@@ -369,7 +358,7 @@ describe("the merger (mergeFindings)", () => {
 		assert.ok(alpha.aliases.includes("Alpha feel"), "the session's phrasing joins the aliases");
 		assert.ok(alpha.aliases.includes("alpha intuition"));
 		assert.ok(
-			alpha.provenance.some((pt) => pt.chatId === chat2.id && pt.messageId === "a1"),
+			alpha.provenance.some((pt) => pt.chatId === chat2.id),
 			"provenance unions",
 		);
 		// The mapping records the verdict: (chat2, q1) → the linked node.
@@ -391,8 +380,6 @@ describe("the merger (mergeFindings)", () => {
 						name: "Vim registers",
 						statement: "Vim registers map single keys to editing commands.",
 						aliases: [],
-						messages: [],
-						evidence: ["fixture evidence"],
 					},
 				],
 				edges: [],
@@ -421,14 +408,12 @@ describe("the merger (mergeFindings)", () => {
 					id: `p${i + 1}`,
 					name: `Cluster node ${i + 1}`,
 					statement,
-					aliases: [],
-					messages: [`cluster_msg_${i + 1}`], // birth provenance — the message anchors
-					evidence: ["fixture evidence"],
+					aliases: [], // birth provenance — the message anchors
 				})),
 				edges: [
-					{ from: "p1", to: "p2", type: "follows", messages: [], evidence: ["e"] },
-					{ from: "p2", to: "p3", type: "follows", messages: [], evidence: ["e"] },
-					{ from: "p1", to: "p3", type: "follows", messages: [], evidence: ["e"] },
+					{ from: "p1", to: "p2", type: "follows", },
+					{ from: "p2", to: "p3", type: "follows", },
+					{ from: "p1", to: "p3", type: "follows", },
 				],
 			},
 			graphDeps,
@@ -477,8 +462,6 @@ describe("the merger (mergeFindings)", () => {
 						name: "Alpha",
 						statement: "Alpha intuition", // IDENTICAL to the existing node's statement
 						aliases: [],
-						messages: ["u1"],
-						evidence: ["fixture evidence"],
 						// no linkTo — identity is the auditor's call alone
 					},
 				],
@@ -502,9 +485,7 @@ describe("the merger (mergeFindings)", () => {
 						id: "p1",
 						name: "Vim normal mode",
 						statement: "Vim starts in Normal mode, where keys are commands.",
-						aliases: [],
-						messages: [], // notes carry no message anchors — the file is the provenance
-						evidence: ["fixture evidence"],
+						aliases: [], // notes carry no message anchors — the file is the provenance
 					},
 				],
 				edges: [],
@@ -515,7 +496,7 @@ describe("the merger (mergeFindings)", () => {
 		assert.deepEqual(stats, { nodes: 1, edges: 0 });

 		const node = getGraph(graphStore).nodes.find((n) => n.statement.startsWith("Vim starts"))!;
-		assert.deepEqual(node.provenance, [], "no message anchors — the mapping line carries the source");
+		assert.deepEqual(node.provenance, [{ chatId: "src_notesfixture", messageId: "" }], "one chat-level point — the source key, no message anchors");
 		assert.deepEqual(graphStore.readMappings().find((m) => m.chatId === "src_notesfixture"), {
 			chatId: "src_notesfixture",
 			sourceId: "p1",
@@ -607,10 +588,10 @@ describe("session deletion (deleteChatTraces + deleteChatSession, wave 6)", () =
 		].flat());
 		const auditA: AuditResult = {
 			propositions: [
-				{ id: "p1", name: "Shared core", statement: "Shared core statement", aliases: ["shared feel"], messages: ["du1"], evidence: ["e"] },
-				{ id: "p2", name: "Solo fact", statement: "Solo fact statement", aliases: ["solo feel"], messages: ["du1"], evidence: ["e"] },
+				{ id: "p1", name: "Shared core", statement: "Shared core statement", aliases: ["shared feel"], },
+				{ id: "p2", name: "Solo fact", statement: "Solo fact statement", aliases: ["solo feel"], },
 			],
-			edges: [{ from: "p1", to: "p2", type: "follows", messages: ["du1"], evidence: ["e"] }],
+			edges: [{ from: "p1", to: "p2", type: "follows", }],
 		};
 		mergeFindings(dA.id, auditA, graphDeps, sessionSource(dA.id));
 		// Resolve through THIS session's mappings — the shared ledger holds
@@ -633,10 +614,10 @@ describe("session deletion (deleteChatTraces + deleteChatSession, wave 6)", () =
 		].flat());
 		const auditB: AuditResult = {
 			propositions: [
-				{ id: "q1", name: "Shared again", statement: "Shared core, rephrased", aliases: ["B phrasing"], messages: ["eu1"], evidence: ["e"], linkTo: xId },
-				{ id: "q2", name: "B extra", statement: "B extra statement", aliases: ["b extra feel"], messages: ["eu1"], evidence: ["e"] },
+				{ id: "q1", name: "Shared again", statement: "Shared core, rephrased", aliases: ["B phrasing"], linkTo: xId },
+				{ id: "q2", name: "B extra", statement: "B extra statement", aliases: ["b extra feel"], },
 			],
-			edges: [{ from: "q1", to: "q2", type: "follows", messages: ["eu1"], evidence: ["e"] }],
+			edges: [{ from: "q1", to: "q2", type: "follows", }],
 		};
 		mergeFindings(dB.id, auditB, graphDeps, sessionSource(dB.id));
 		const zId = graphStore.readMappings().find((m) => m.chatId === dB.id && m.sourceId === "q2")!.graphNodeId;
@@ -673,7 +654,7 @@ describe("session deletion (deleteChatTraces + deleteChatSession, wave 6)", () =
 		assert.equal(x.name, "Shared core");
 		assert.ok(!x.aliases.includes("shared feel"), "the deleted creator's phrasings are stripped (owner ruling)");
 		assert.ok(x.aliases.includes("B phrasing") && x.aliases.includes("Shared core, rephrased"));
-		assert.deepEqual(x.provenance, [{ chatId: dB.id, messageId: "eu1" }], "only dB's provenance remains");
+		assert.deepEqual(x.provenance, [{ chatId: dB.id, messageId: "" }], "only dB's provenance remains");
 		// Y (the orphan) is gone entirely.
 		assert.equal(graph.nodes.some((n) => n.id === yId), false);
 		assert.equal(graphStore.readNodes().some((n) => n.id === yId), false);
@@ -681,7 +662,7 @@ describe("session deletion (deleteChatTraces + deleteChatSession, wave 6)", () =
 		assert.equal(JSON.stringify(graphStore.readNodes().find((n) => n.id === zId)), zLineBefore);
 		// Edges: dA's assertion X→Y died; dB's X→Z survived.
 		const scoped = scopedGraph([dA.id, dB.id]);
-		assert.deepEqual(scoped.edges, [{ from: xId, to: zId, type: "follows", provenance: [{ chatId: dB.id, messageId: "eu1" }] }]);
+		assert.deepEqual(scoped.edges, [{ from: xId, to: zId, type: "follows", provenance: [{ chatId: dB.id, messageId: "" }] }]);
 		// Mappings: dA's audit trail died; dB's stays.
 		const mappings = graphStore.readMappings().filter((m) => [dA.id, dB.id].includes(m.chatId));
 		assert.deepEqual(mappings, [{ chatId: dB.id, sourceId: "q1", graphNodeId: xId }, { chatId: dB.id, sourceId: "q2", graphNodeId: zId }]);