mirror of
https://github.com/paperclipai/paperclip.git
synced 2026-11-12 06:10:44 +01:00
Publish isolated native qualification manifest and exact admissions
Co-Authored-By: Paperclip <noreply@paperclip.ing>
This commit is contained in:
DottaandPaperclip committed
1 parent
2d6fc5eb03
commit
93f6fc9356
6 files changed
+1002
No files matched your search
@@ -0,0 +1,258 @@
|
||||
{
|
||||
"schema": "paperclip.native-completion-preflight.v1",
|
||||
"sourceSha": "459455acb11a012a97ad1b4afb77a1dc024a88bc",
|
||||
"sourceFingerprint": "4b49ad1bd8acaabc4a69700729295fe31395d91aa19fa650e8066988910cc681",
|
||||
"fixtureFingerprint": "38ae4fe73494db231ed50d0c2336762a290cc17ebc54f258bd94f5a0b873ec09",
|
||||
"manifestFingerprint": "b92f3ab7d561188bc09214433b264f2e7810e8ccdc237f68caa82e1aecc1811e",
|
||||
"variant": "historical",
|
||||
"baseSha": "59c07ede72dc08b8aba149a01cc11e0b7a204621",
|
||||
"archiveSha": "9138f570c341c251a5727c32d6615ce238bc8e03",
|
||||
"sourceErrors": [],
|
||||
"immutable": true,
|
||||
"layering": true,
|
||||
"measuredAt": "2026-10-03T02:53:32.494Z",
|
||||
"providerCalls": 0,
|
||||
"live": "not_run",
|
||||
"expectedCells": 6,
|
||||
"expectedTurns": 6,
|
||||
"maximumAttemptsPerCell": 1,
|
||||
"setup": {
|
||||
"passed": true,
|
||||
"sdkExitCode": 0,
|
||||
"runnerTypeScriptExitCode": 0,
|
||||
"evidence": [
|
||||
{
|
||||
"file": "setup-sdk.txt",
|
||||
"sha256": "34d8ecdeb2419da9426e71ac595501ee653545aa2e41de92eabf14420221da86"
|
||||
},
|
||||
{
|
||||
"file": "setup-runner-typescript.txt",
|
||||
"sha256": "73d1d585d121f9d912f3ceb28d5dee52269630bd89afc705d01fcd4103db41ec"
|
||||
}
|
||||
],
|
||||
"buildOutputFingerprint": "1c1e98bce110b2cdbe30f6c2d21c0a0afc8c51b5ae4c7f9bd11580ff9dd62262"
|
||||
},
|
||||
"gates": [
|
||||
{
|
||||
"id": "NC-schemas",
|
||||
"name": "Variant-native completion schemas",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/contracts/completion-result.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "distinguishes user-facing answer content",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "propagates answer and wait descriptions",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "requires a reason code for verification that was not run",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 13,
|
||||
"passedTests": 13,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-schemas.json",
|
||||
"sha256": "5fbb694c27781d5d3f2bda026130da5a386ac70d38d9bd63e5ecd901d309778f"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-tools",
|
||||
"name": "Native tool delivery and retained provider catalogs",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/drivers/codex/codex-app-server-driver.lifecycle.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/runner-tool-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/opencode/mcp-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/live/runnerd-codex-transport.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "runner semantic MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "OpenCode MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on task recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on prepared recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on direct recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "includes ACPX terminal tools in the authenticated bridge catalog",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 222,
|
||||
"passedTests": 7,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 215,
|
||||
"evidence": {
|
||||
"file": "NC-tools.json",
|
||||
"sha256": "02c32abf655f22c7a046271f37be749a300113e3f9056e088a231d8ed0bd872a"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-resume",
|
||||
"name": "Variant-native fingerprint and checkpoint refresh",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "server/src/services/native-runtime/native-session-resume.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "refreshes retained 'completion descriptions'",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "refreshes retained 'task-bound human-input description'",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 44,
|
||||
"passedTests": 2,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 42,
|
||||
"evidence": {
|
||||
"file": "NC-resume.json",
|
||||
"sha256": "649f0d367c7608c0885be2dd9817eec2a9805476fa7cb0d7a0092e8b95318cfc"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-eval",
|
||||
"name": "Independent native oracle, defaults, admission and attempt policy",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-scoring.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-defaults.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-admission.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/automatic-retry.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-evidence.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-flow-logs.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "allows toolchain paths and excludes every present or future credential",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains credential-free prerequisites inside the exact campaign root",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first transient_infrastructure failure without a second attempt",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first provider_variance failure without a second attempt",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 99,
|
||||
"passedTests": 99,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-eval.json",
|
||||
"sha256": "c9ec21cfad75d5afbfd95049c3490ac78be69b793439db32c12b8bd16568e03c"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-node",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-node.txt",
|
||||
"sha256": "2720099b4b95dbace32017d8026b4e94c83aa67718ca0550a682909cbfa6c26d"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-typecheck",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-typecheck.txt",
|
||||
"sha256": "01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-discovery",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-discovery.txt",
|
||||
"sha256": "3097ab7907c3b49d43920fab50d0b180b70cb6895c0bd1a58f72fd740fec6567"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-manifest",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-manifest.txt",
|
||||
"sha256": "7f5d044db6a7813857e842095c5e3aa749a98570a96ff9939483de9452322502"
|
||||
}
|
||||
}
|
||||
],
|
||||
"passed": true
|
||||
}
|
||||
@@ -0,0 +1,282 @@
|
||||
{
|
||||
"schema": "paperclip.native-completion-preflight.v1",
|
||||
"sourceSha": "e18c2cf9e96a4d31acb6d03ce918dfe223107d3f",
|
||||
"sourceFingerprint": "29791687f245dde6442064b855b2a76f0be8c1ec33635ac174688ae06732d33e",
|
||||
"fixtureFingerprint": "38ae4fe73494db231ed50d0c2336762a290cc17ebc54f258bd94f5a0b873ec09",
|
||||
"manifestFingerprint": "b92f3ab7d561188bc09214433b264f2e7810e8ccdc237f68caa82e1aecc1811e",
|
||||
"variant": "candidate",
|
||||
"baseSha": "59c07ede72dc08b8aba149a01cc11e0b7a204621",
|
||||
"archiveSha": "9138f570c341c251a5727c32d6615ce238bc8e03",
|
||||
"sourceErrors": [],
|
||||
"immutable": true,
|
||||
"layering": true,
|
||||
"measuredAt": "2026-10-03T02:53:32.991Z",
|
||||
"providerCalls": 0,
|
||||
"live": "not_run",
|
||||
"expectedCells": 6,
|
||||
"expectedTurns": 6,
|
||||
"maximumAttemptsPerCell": 1,
|
||||
"setup": {
|
||||
"passed": true,
|
||||
"sdkExitCode": 0,
|
||||
"runnerTypeScriptExitCode": 0,
|
||||
"evidence": [
|
||||
{
|
||||
"file": "setup-sdk.txt",
|
||||
"sha256": "01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b"
|
||||
},
|
||||
{
|
||||
"file": "setup-runner-typescript.txt",
|
||||
"sha256": "34e7d309170c1981087ead95522b7dda69dffc74a8fd9ad68910e10f568aaa87"
|
||||
}
|
||||
],
|
||||
"buildOutputFingerprint": "b98c7815799b3b663b6d9082325212868cfbf59e74c94be4d6a8ce09b751d443"
|
||||
},
|
||||
"gates": [
|
||||
{
|
||||
"id": "NC-schemas",
|
||||
"name": "Variant-native completion schemas",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/contracts/completion-result.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "distinguishes user-facing answer content",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "propagates answer and wait descriptions",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "requires a reason code for verification that was not run",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 13,
|
||||
"passedTests": 13,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-schemas.json",
|
||||
"sha256": "3f7c907d2a72c25e85d7944831becf346055bbb63e5be7b0264bb01c6dcd0dd4"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-tools",
|
||||
"name": "Native tool delivery and retained provider catalogs",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/drivers/codex/codex-app-server-driver.lifecycle.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/runner-tool-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/opencode/mcp-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/live/runnerd-codex-transport.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "runner semantic MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "OpenCode MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on task recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on prepared recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on direct recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "includes ACPX terminal tools in the authenticated bridge catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native codex tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native opencode tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native claude_managed tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native aws_agentcore tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native acpx tool catalog",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 226,
|
||||
"passedTests": 11,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 215,
|
||||
"evidence": {
|
||||
"file": "NC-tools.json",
|
||||
"sha256": "2d55bd1da955c203fb9026700be5c6df407eb57fedea603900819831b49e0677"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-resume",
|
||||
"name": "Variant-native fingerprint and checkpoint refresh",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "server/src/services/native-runtime/native-session-resume.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "refreshes retained 'completion descriptions'",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "refreshes retained 'task-bound human-input description'",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "native completion tool guidance",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 45,
|
||||
"passedTests": 3,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 42,
|
||||
"evidence": {
|
||||
"file": "NC-resume.json",
|
||||
"sha256": "4f10df44bba11f4d55706c8b73f3d8b07d8e7a5c965d9b8473d95f204ffbc904"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-eval",
|
||||
"name": "Independent native oracle, defaults, admission and attempt policy",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-scoring.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-defaults.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-admission.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/automatic-retry.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-evidence.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-flow-logs.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "allows toolchain paths and excludes every present or future credential",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains credential-free prerequisites inside the exact campaign root",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first transient_infrastructure failure without a second attempt",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first provider_variance failure without a second attempt",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 99,
|
||||
"passedTests": 99,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-eval.json",
|
||||
"sha256": "049b0f5d3b2024d6fb1704f574566298175c9a00831118dd3a09b4590a0a2304"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-node",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-node.txt",
|
||||
"sha256": "0a9c776c6075c8e800461f97a1fb3e32cedd89185104cf0e8e79a5d157f82328"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-typecheck",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-typecheck.txt",
|
||||
"sha256": "01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-discovery",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-discovery.txt",
|
||||
"sha256": "3097ab7907c3b49d43920fab50d0b180b70cb6895c0bd1a58f72fd740fec6567"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-manifest",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-manifest.txt",
|
||||
"sha256": "7f5d044db6a7813857e842095c5e3aa749a98570a96ff9939483de9452322502"
|
||||
}
|
||||
}
|
||||
],
|
||||
"passed": true
|
||||
}
|
||||
+282
@@ -0,0 +1,282 @@
|
||||
{
|
||||
"schema": "paperclip.native-completion-preflight.v1",
|
||||
"sourceSha": "b8a6bf950a2072ae8e6ba2ecd495b40978ba9961",
|
||||
"sourceFingerprint": "44b3fe315b0c0fcd343b2e64120a703ca02b73beca7902e79b1a837b122d2d7b",
|
||||
"fixtureFingerprint": "687a5b5bc562b6b31b781436b4adcabaf52109ce30bce4897c561244e9884246",
|
||||
"manifestFingerprint": "b92f3ab7d561188bc09214433b264f2e7810e8ccdc237f68caa82e1aecc1811e",
|
||||
"variant": "candidate",
|
||||
"baseSha": "59c07ede72dc08b8aba149a01cc11e0b7a204621",
|
||||
"archiveSha": "9138f570c341c251a5727c32d6615ce238bc8e03",
|
||||
"sourceErrors": [],
|
||||
"immutable": true,
|
||||
"layering": true,
|
||||
"measuredAt": "2026-10-03T02:50:39.618Z",
|
||||
"providerCalls": 0,
|
||||
"live": "not_run",
|
||||
"expectedCells": 6,
|
||||
"expectedTurns": 6,
|
||||
"maximumAttemptsPerCell": 1,
|
||||
"setup": {
|
||||
"passed": true,
|
||||
"sdkExitCode": 0,
|
||||
"runnerTypeScriptExitCode": 0,
|
||||
"evidence": [
|
||||
{
|
||||
"file": "setup-sdk.txt",
|
||||
"sha256": "34d8ecdeb2419da9426e71ac595501ee653545aa2e41de92eabf14420221da86"
|
||||
},
|
||||
{
|
||||
"file": "setup-runner-typescript.txt",
|
||||
"sha256": "34e7d309170c1981087ead95522b7dda69dffc74a8fd9ad68910e10f568aaa87"
|
||||
}
|
||||
],
|
||||
"buildOutputFingerprint": "b98c7815799b3b663b6d9082325212868cfbf59e74c94be4d6a8ce09b751d443"
|
||||
},
|
||||
"gates": [
|
||||
{
|
||||
"id": "NC-schemas",
|
||||
"name": "Variant-native completion schemas",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/contracts/completion-result.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "distinguishes user-facing answer content",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "propagates answer and wait descriptions",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "requires a reason code for verification that was not run",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 13,
|
||||
"passedTests": 13,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-schemas.json",
|
||||
"sha256": "2c69600d93f3a69797e27ff2928f0d1668eb59443b1dff7a9255b2881ba0e305"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-tools",
|
||||
"name": "Native tool delivery and retained provider catalogs",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "src/drivers/codex/codex-app-server-driver.lifecycle.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/runner-tool-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/drivers/opencode/mcp-bridge.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "src/live/runnerd-codex-transport.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "runner semantic MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "OpenCode MCP bridge binds to loopback",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on task recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on prepared recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves stock Codex instructions on direct recovery",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "includes ACPX terminal tools in the authenticated bridge catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native codex tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native opencode tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native claude_managed tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native aws_agentcore tool catalog",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "preserves answer and internal wait descriptions in the serialized native acpx tool catalog",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 226,
|
||||
"passedTests": 11,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 215,
|
||||
"evidence": {
|
||||
"file": "NC-tools.json",
|
||||
"sha256": "e60ff6fdb64b302136c323a27f6b88d49284c13a313b349c40c039519280bcab"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-resume",
|
||||
"name": "Variant-native fingerprint and checkpoint refresh",
|
||||
"passed": false,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "server/src/services/native-runtime/native-session-resume.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "refreshes retained completion descriptions",
|
||||
"passed": false
|
||||
},
|
||||
{
|
||||
"name": "refreshes retained task-bound human-input description",
|
||||
"passed": false
|
||||
},
|
||||
{
|
||||
"name": "native completion tool guidance",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 45,
|
||||
"passedTests": 3,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 42,
|
||||
"evidence": {
|
||||
"file": "NC-resume.json",
|
||||
"sha256": "416177a10dc3daba2824656d15439358cc057f00e65c0478eb65b1b30cb30785"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-eval",
|
||||
"name": "Independent native oracle, defaults, admission and attempt policy",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"files": [
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-scoring.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-defaults.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/native-completion-admission.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/automatic-retry.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-evidence.test.ts",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"file": "tests/runner-e2e/context-integrity-flow-logs.test.ts",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"requirements": [
|
||||
{
|
||||
"name": "allows toolchain paths and excludes every present or future credential",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains credential-free prerequisites inside the exact campaign root",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first transient_infrastructure failure without a second attempt",
|
||||
"passed": true
|
||||
},
|
||||
{
|
||||
"name": "retains the first provider_variance failure without a second attempt",
|
||||
"passed": true
|
||||
}
|
||||
],
|
||||
"total": 99,
|
||||
"passedTests": 99,
|
||||
"failedTests": 0,
|
||||
"pendingTests": 0,
|
||||
"evidence": {
|
||||
"file": "NC-eval.json",
|
||||
"sha256": "1bed25972c4332dbaa81e6911a550ca01f66a0c710c8a20dc44f984b73dae2e7"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-node",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-node.txt",
|
||||
"sha256": "bf411b588d0b0ab413ba647aacdef6fb7866a4ed81055ef5883c0bb2494b6b1a"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-typecheck",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-typecheck.txt",
|
||||
"sha256": "01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-discovery",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-discovery.txt",
|
||||
"sha256": "3097ab7907c3b49d43920fab50d0b180b70cb6895c0bd1a58f72fd740fec6567"
|
||||
}
|
||||
},
|
||||
{
|
||||
"id": "NC-manifest",
|
||||
"passed": true,
|
||||
"exitCode": 0,
|
||||
"evidence": {
|
||||
"file": "NC-manifest.txt",
|
||||
"sha256": "7f5d044db6a7813857e842095c5e3aa749a98570a96ff9939483de9452322502"
|
||||
}
|
||||
}
|
||||
],
|
||||
"passed": false
|
||||
}
|
||||
@@ -0,0 +1,148 @@
|
||||
{
|
||||
"schema": "paperclip.native-completion-master-comparison.v1",
|
||||
"baseSha": "59c07ede72dc08b8aba149a01cc11e0b7a204621",
|
||||
"archiveSha": "9138f570c341c251a5727c32d6615ce238bc8e03",
|
||||
"admission": {
|
||||
"candidate": {
|
||||
"sourceSha": "e18c2cf9e96a4d31acb6d03ce918dfe223107d3f",
|
||||
"variant": "candidate",
|
||||
"sourceFingerprint": "29791687f245dde6442064b855b2a76f0be8c1ec33635ac174688ae06732d33e",
|
||||
"fixtureFingerprint": "38ae4fe73494db231ed50d0c2336762a290cc17ebc54f258bd94f5a0b873ec09",
|
||||
"manifestFingerprint": "b92f3ab7d561188bc09214433b264f2e7810e8ccdc237f68caa82e1aecc1811e",
|
||||
"receiptSha256": "74b5d3b99b059ceda2f572154595038e4c10dbc2af269e018eefc8cb59d60d38",
|
||||
"passed": true,
|
||||
"providerCalls": 0
|
||||
},
|
||||
"baseline": {
|
||||
"sourceSha": "459455acb11a012a97ad1b4afb77a1dc024a88bc",
|
||||
"variant": "historical",
|
||||
"sourceFingerprint": "4b49ad1bd8acaabc4a69700729295fe31395d91aa19fa650e8066988910cc681",
|
||||
"fixtureFingerprint": "38ae4fe73494db231ed50d0c2336762a290cc17ebc54f258bd94f5a0b873ec09",
|
||||
"manifestFingerprint": "b92f3ab7d561188bc09214433b264f2e7810e8ccdc237f68caa82e1aecc1811e",
|
||||
"receiptSha256": "affc25f42cca90c66cbe1365fbfd47d33657e59228730de99fc6194871aed1f8",
|
||||
"passed": true,
|
||||
"providerCalls": 0
|
||||
}
|
||||
},
|
||||
"targetBranches": {
|
||||
"candidate": "codex/native-completion-current-master-candidate",
|
||||
"baseline": "codex/native-completion-current-master-baseline"
|
||||
},
|
||||
"workflow": ".github/workflows/runner-full-stack-e2e.yml",
|
||||
"workflowDispatchRef": "master",
|
||||
"maximumCells": 12,
|
||||
"expectedProviderTurns": 12,
|
||||
"maximumAttemptsPerCell": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"budgetMonthlyCentsPerCompanyAndAgent": 1000,
|
||||
"cellsPerVariant": [
|
||||
{
|
||||
"executionId": "native-completion.runner-codex.local.assigned-skill-explicit-invocation",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-codex",
|
||||
"provider": "codex",
|
||||
"model": "gpt-5.6-sol",
|
||||
"credential": "OPENAI_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "assigned-skill-explicit-invocation",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 720000,
|
||||
"modelQualification": {
|
||||
"source": "adapter_constant",
|
||||
"qualificationId": "codex:qualified-model"
|
||||
}
|
||||
},
|
||||
{
|
||||
"executionId": "native-completion.runner-codex.local.native-blocked-report",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-codex",
|
||||
"provider": "codex",
|
||||
"model": "gpt-5.6-sol",
|
||||
"credential": "OPENAI_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "native-blocked-report",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 300000,
|
||||
"modelQualification": {
|
||||
"source": "adapter_constant",
|
||||
"qualificationId": "codex:qualified-model"
|
||||
}
|
||||
},
|
||||
{
|
||||
"executionId": "native-completion.runner-opencode.local.assigned-skill-explicit-invocation",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-opencode",
|
||||
"provider": "opencode",
|
||||
"model": "openrouter/deepseek/deepseek-v4-flash-0731",
|
||||
"credential": "OPENROUTER_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "assigned-skill-explicit-invocation",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 720000,
|
||||
"modelQualification": {
|
||||
"source": "adapter_constant",
|
||||
"qualificationId": "opencode:qualified-model"
|
||||
}
|
||||
},
|
||||
{
|
||||
"executionId": "native-completion.runner-opencode.local.native-blocked-report",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-opencode",
|
||||
"provider": "opencode",
|
||||
"model": "openrouter/deepseek/deepseek-v4-flash-0731",
|
||||
"credential": "OPENROUTER_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "native-blocked-report",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 300000,
|
||||
"modelQualification": {
|
||||
"source": "adapter_constant",
|
||||
"qualificationId": "opencode:qualified-model"
|
||||
}
|
||||
},
|
||||
{
|
||||
"executionId": "native-completion.runner-acpx-claude.local.assigned-skill-explicit-invocation",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-acpx-claude",
|
||||
"provider": "acpx",
|
||||
"model": "claude-sonnet-5",
|
||||
"credential": "ANTHROPIC_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "assigned-skill-explicit-invocation",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 720000,
|
||||
"modelQualification": {
|
||||
"source": "qualified_runner_profile",
|
||||
"qualificationId": "acpx:claude"
|
||||
}
|
||||
},
|
||||
{
|
||||
"executionId": "native-completion.runner-acpx-claude.local.native-blocked-report",
|
||||
"suiteDefinitionHash": "715cd439b1a395980f7adaa3893f69aee4da7e53b2b7f87a9efd6d2c33ec605e",
|
||||
"profileId": "runner-acpx-claude",
|
||||
"provider": "acpx",
|
||||
"model": "claude-sonnet-5",
|
||||
"credential": "ANTHROPIC_API_KEY",
|
||||
"environment": "local",
|
||||
"caseId": "native-blocked-report",
|
||||
"expectedRunCount": 1,
|
||||
"automaticRetryPolicy": "single_attempt",
|
||||
"deadlineMs": 300000,
|
||||
"modelQualification": {
|
||||
"source": "qualified_runner_profile",
|
||||
"qualificationId": "acpx:claude"
|
||||
}
|
||||
}
|
||||
],
|
||||
"fixtureIdentityVerified": true,
|
||||
"fixedMasterContextVerified": true,
|
||||
"onlyDeclaredElevenNativeSourcesDiffer": true,
|
||||
"providerCallsSoFar": 0,
|
||||
"liveStatus": "pending",
|
||||
"workflowRuns": {}
|
||||
}
|
||||
@@ -0,0 +1,27 @@
|
||||
# Native completion guidance: master-context comparison
|
||||
|
||||
**TL;DR:** Live non-regression is pending: zero graded pairs, six pending pairs. Both exact-source credential-free admissions pass. This comparison measures native finish/block descriptions on unchanged master defaults; it does not measure removal of native fixed prompts or the reduced manuals in #14948.
|
||||
|
||||
Candidate `e18c2cf9e96a4d31acb6d03ce918dfe223107d3f` and historical-description baseline `459455acb11a012a97ad1b4afb77a1dc024a88bc` share master `59c07ede72dc08b8aba149a01cc11e0b7a204621`. Their executable fixture fingerprint is `38ae4fe73494db231ed50d0c2336762a290cc17ebc54f258bd94f5a0b873ec09`. Only the declared five native production files and six matching unit-test files differ. Default manual, shared prompts, operational skill, runtime context, profile/model/auth/permission configuration and manifest bytes remain fixed.
|
||||
|
||||
The [frozen manifest](2026-10-02-native-completion-master-manifest.json) lists all six IDs, actual configured models, exact source/admission hashes, deadlines and single-attempt policy. Candidate and baseline will run concurrently through the trusted master workflow on distinct immutable target branches. Scope is twelve cells/twelve expected provider turns, with no automatic retries and 1,000-cent company/agent hard stops per cell. No providers have run at this publication.
|
||||
|
||||
| Profile | Preserved assigned-skill durable-document completion | Whole-task blocker |
|
||||
|---|---|---|
|
||||
| Codex | Pending pair | Pending pair |
|
||||
| ACPX Claude | Pending pair | Pending pair |
|
||||
| OpenCode | Pending pair | Pending pair |
|
||||
|
||||
## Admission and retained setup failures
|
||||
|
||||
[Candidate receipt](2026-10-02-native-completion-master-admission/candidate.json): 126 selected TypeScript assertions pass. [Historical receipt](2026-10-02-native-completion-master-admission/baseline.json): 121 selected assertions pass. Both pass 73 Node calibrations, E2E typecheck, manifest checks, six-cell discovery and source/build receipt verification. The 257 unrelated assertions explicitly skipped in each variant are not counted as coverage. The candidate's additional five assertions cover new daemon catalogs and v13 rotation; historical unit tests validate historical descriptions instead. Independent focused integration passes 130 assertions; ten browser blocker calibrations pass without providers.
|
||||
|
||||
[Initial candidate admission](2026-10-02-native-completion-master-admission/failed-candidate-b8a6bf950.json) is preserved. All executable tests passed, but the gate's required-name matcher omitted Vitest's quotes in two parameterized resume test names, so admission failed closed. The corrected matcher is identical in both final variants; behavioral assertions were unchanged. Baseline sandboxed offline dependency setup initially reported a cache miss; authorized cache access completed the same frozen-lock install. No dependency/lock changes or provider calls occurred in either setup attempt.
|
||||
|
||||
## Evidence boundary
|
||||
|
||||
The completion case preserves the original assigned-skill prompt, installed marker-only output skill, explicit invocation and independently saved issue document oracle. Production-default hires are verified through the public served bundle before execution. The blocker case preserves the exact owner/action/scope and independently checks durable Blocked status, explanatory visible reply, no additional domain work and no shell/process/file work.
|
||||
|
||||
Both require exactly one succeeded native run bound to the issue, complete public events, a matched finishing-tool start/result, actual provider final after that result, no new invocation after proposal, and its persisted run-attributed reply. Authoritative control-plane acceptance/terminal are independent of final timing. Observable ordering does not prove feedback consumption. A single pair cannot establish general coding-quality equivalence. Missing or failed evidence stays explicit.
|
||||
|
||||
The [earlier report](https://github.com/paperclipai/paperclip/blob/9138f570c341c251a5727c32d6615ce238bc8e03/doc/plans/2026-10-02-native-completion-live-comparison.md) and archived source remain unchanged. Its original verdicts, grader limitations and provider-free replays are historical evidence from the reduced-manual context, not measurements of these new frozen refs.
|
||||
@@ -494,3 +494,8 @@ results, remaining exceptions, and follow-ups here before checking it off.
|
||||
- F26: Read-only native hiring carrier audit confirms deliberate source-target loss in both frozen variants and an ACPX full-result hash/count recomputation from a roughly 4 KiB diagnostic preview. This does not establish provider-visible truncation or a prompt regression. Closed managed-source receipts and truthful full-result metadata are proposals only; no production carrier changes or retrospective source-read regrade. [Inspectable audit](2026-10-02-hiring-read-evidence-carrier-audit.md).
|
||||
|
||||
- F27: Native hiring completion actions carry internally paired native API IDs, but ACPX host request IDs differ from canonical provider execution IDs with no established mapping. Strict v3 accounting does not join by name/order/count; both ACPX action assessments remain uncomparable despite only successful native GET receipts being observed. Codex exact IDs verify both variants. Keep this action-attribution gap separate from source-read F25/F26, and do not weaken the carrier/oracle or rerun providers.
|
||||
|
||||
|
||||
### Native completion isolation: current-master qualification
|
||||
|
||||
PR #14961 is now independent of #14948 and based on master. Candidate `e18c2cf9e96a4d31acb6d03ce918dfe223107d3f` and historical baseline `459455acb11a012a97ad1b4afb77a1dc024a88bc` pass exact-source native-only admission with identical fixture/context receipts. The original assigned-skill/document journey is preserved; corrected blocker grading and enforced single attempts cover six paired native cells. Live results remain pending. [Manifest and current report](2026-10-02-native-completion-master-qualification.md) retain setup failures and the earlier native report separately. #14948's behavioral hold and #15007's attribution limits remain unchanged. No merge is authorized for #14961 yet.
|
||||
Reference in new issue
Block a user