From aa4b88b169849645038eeb6d305b533767dad77d Mon Sep 17 00:00:00 2001
From: Dotta <bippadotta@protonmail.com>
Date: Mon, 28 Sep 2026 14:46:49 -0500
Subject: [PATCH] fix(runner): include live completion contract in model
 instructions

Co-Authored-By: Paperclip <noreply@paperclip.ing>
---
 .../src/live/live-session.test.ts                 |  8 +++++++-
 .../paperclip-runner/src/live/live-session.ts     | 15 +++++++++++----
 2 files changed, 18 insertions(+), 5 deletions(-)

diff --git a/packages/paperclip-runner/src/live/live-session.test.ts b/packages/paperclip-runner/src/live/live-session.test.ts
index f565840c151..f1d7a1d37ce 100644
--- a/packages/paperclip-runner/src/live/live-session.test.ts
+++ b/packages/paperclip-runner/src/live/live-session.test.ts
@@ -907,6 +907,8 @@ describe("Capability live runnerd and Codex session", () => {
     expect(names.filter((name) => name === "paperclip_finish")).toHaveLength(1);
     expect(names.filter((name) => name === "paperclip_block")).toHaveLength(1);
     expect(names).not.toContain("create_task");
+    expect(opened.params.baseInstructions).toContain('"revision":"paperclip-capability-live-v1"');
+    expect(opened.params.baseInstructions).toContain('"criterionIds":["objective"]');
     await first.suspend();
     const restoredService = new CapabilityLiveSessionService({ store, transportFactory: factory });
     const restored = await restoredService.restore(first.id);
@@ -914,6 +916,8 @@ describe("Capability live runnerd and Codex session", () => {
     expect(authority[1]?.resumeDynamicTools?.map((tool) => tool.name)).toEqual(
       names.filter((name) => name !== "paperclip_finish" && name !== "paperclip_block"),
     );
+    const resumed = state.transports[1]!.requests.find((request) => request.method === "thread/resume")!;
+    expect(resumed.params.baseInstructions).toBe(opened.params.baseInstructions);
     await restoredService.shutdown(restored.id);
     await firstService.shutdown(first.id);
   });
@@ -972,9 +976,11 @@ describe("Capability live runnerd and Codex session", () => {
       state.transports[0]?.requests.find(
         (request) => request.method === "thread/start",
       )?.params.baseInstructions,
-    ).toBe(
+    ).toContain(
       "Native instructions\n\nRead-only instruction sibling root: /runtime/instructions",
     );
+    expect(state.transports[0]?.requests.find((request) => request.method === "thread/start")?.params.baseInstructions)
+      .toContain('Native completion report contract: {"revision":"paperclip-capability-live-v1","criterionIds":["objective"]}');
     await service.shutdown(session.id);
   });
 
diff --git a/packages/paperclip-runner/src/live/live-session.ts b/packages/paperclip-runner/src/live/live-session.ts
index 35339216162..ecf16f24320 100644
--- a/packages/paperclip-runner/src/live/live-session.ts
+++ b/packages/paperclip-runner/src/live/live-session.ts
@@ -2346,6 +2346,15 @@ export class CapabilityLiveSession {
       itemId: `item_lab_${identityDigest}`,
     };
     this.#providerRunBinding = providerRunBinding;
+    // PRP run-attach metadata binds server validation, but ACP providers receive
+    // only these system instructions. Make the same immutable contract visible
+    // without expanding the model's separately governed mock task authority.
+    const baseInstructions = [
+      this.#transportOptions.baseInstructions ?? LIVE_BASE_INSTRUCTIONS,
+      "",
+      "Native completion report contract: " + JSON.stringify(LIVE_COMPLETION_CONTRACT),
+      "For paperclip_finish or paperclip_block, use that exact contract revision and criterion IDs. These tools report the current provider run result; they do not change mock task state. Mock task mutations still require separately authorized semantic tools and an explicit user request.",
+    ].join("\n");
     const authorizedTools = this.#dispatcher.listTools(this.#authority.runId);
     const tools = (this.#config.toolExposure ?? "eager") === "lazy"
       ? authorizedTools.filter((tool) => tool.annotations.exposure === "always")
@@ -2464,8 +2473,7 @@ export class CapabilityLiveSession {
         config: createSkilllessCodexThreadConfig(this.#config.workingDirectory),
         permissions: CODEX_PERMISSION_PROFILE,
         runtimeWorkspaceRoots: [this.#config.workingDirectory],
-        baseInstructions:
-          this.#transportOptions.baseInstructions ?? LIVE_BASE_INSTRUCTIONS,
+        baseInstructions,
         persistExtendedHistory: true,
       });
       const resumedThread = record(resumed.thread);
@@ -2490,8 +2498,7 @@ export class CapabilityLiveSession {
         permissions: CODEX_PERMISSION_PROFILE,
         runtimeWorkspaceRoots: [this.#config.workingDirectory],
         approvalPolicy: "never",
-        baseInstructions:
-          this.#transportOptions.baseInstructions ?? LIVE_BASE_INSTRUCTIONS,
+        baseInstructions,
         completionContract: LIVE_COMPLETION_CONTRACT,
         dynamicTools: [...semanticTools, ...codexSemanticToolSpecs()],
         experimentalRawEvents: true,