crysk 0.0.0 → 0.3.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/LICENSE +176 -0
- package/README.md +609 -3
- package/bin/pr-hero.js +166 -0
- package/config/models/anthropic.json +5 -0
- package/dist/SKILL-7e3725zn.md +496 -0
- package/dist/SKILL-7q35wcaz.md +66 -0
- package/dist/SKILL-9at22381.md +74 -0
- package/dist/SKILL-ex2hxxh3.md +134 -0
- package/dist/adjudicator-dtg2axnh.md +86 -0
- package/dist/admission-config.example-a9vqmmad.json +9 -0
- package/dist/ci-admission-camfbx2b.md +122 -0
- package/dist/cli.js +42292 -0
- package/dist/consolidate-ac0mnfyj.md +24 -0
- package/dist/deep-review-lifecycle-m698ybjp.md +362 -0
- package/dist/deep-review-logic-a6b1gczf.md +156 -0
- package/dist/deep-review-parity-0ec904db.md +120 -0
- package/dist/deep-review-reliability-qyakbjb9.md +125 -0
- package/dist/deep-review-resilience-3tz8y26e.md +118 -0
- package/dist/index.js +20190 -0
- package/dist/opencode-ci-7b0be64p.md +141 -0
- package/dist/review-refuter-j4e9gezy.md +71 -0
- package/dist/scout-yvp153q7.md +140 -0
- package/dist/summarizer-zca44nxa.md +18 -0
- package/dist/workflow-w1g9szna.yml +151 -0
- package/package.json +81 -4
- package/prompts/consolidate.md +24 -0
- package/prompts/default/PROVENANCE.md +18 -0
- package/prompts/default/deep-review-lifecycle.md +362 -0
- package/prompts/default/deep-review-logic.md +156 -0
- package/prompts/default/deep-review-parity.md +120 -0
- package/prompts/default/deep-review-reliability.md +125 -0
- package/prompts/default/deep-review-resilience.md +118 -0
- package/prompts/default/review-refuter.md +71 -0
- package/prompts/scout.md +140 -0
- package/prompts/summarizer.md +18 -0
- package/skills/crysk/SKILL.md +74 -0
- package/skills/crysk-ci-setup/SKILL.md +134 -0
- package/skills/crysk-ci-setup/assets/admission-config.example.json +9 -0
- package/skills/crysk-ci-setup/assets/workflow.yml +151 -0
- package/skills/crysk-ci-setup/references/ci-admission.md +122 -0
- package/skills/crysk-ci-setup/references/opencode-ci.md +141 -0
- package/skills/crysk-loop/SKILL.md +496 -0
- package/skills/crysk-loop/adjudicator.md +86 -0
- package/skills/crysk-pr/SKILL.md +66 -0
- package/src/agent-env.ts +1084 -0
- package/src/asset-manifest.ts +83 -0
- package/src/assets.d.ts +6 -0
- package/src/assets.ts +183 -0
- package/src/brand-paths.ts +155 -0
- package/src/brand.ts +255 -0
- package/src/ci/admission-ledger.ts +350 -0
- package/src/ci/classify-error.ts +73 -0
- package/src/ci/cli-record.ts +271 -0
- package/src/ci/force-comment.ts +39 -0
- package/src/ci/gates.ts +923 -0
- package/src/ci/outcome.ts +2181 -0
- package/src/ci/override.ts +187 -0
- package/src/ci/publish.ts +1639 -0
- package/src/ci/reporter.ts +626 -0
- package/src/ci/review-admission.ts +901 -0
- package/src/ci/review-levels.ts +310 -0
- package/src/ci/review-policy.ts +18 -0
- package/src/ci/review-risk.ts +184 -0
- package/src/ci/setup.ts +844 -0
- package/src/ci/trigger.ts +1353 -0
- package/src/cli.ts +348 -0
- package/src/commands/activity.ts +67 -0
- package/src/commands/ci-setup-questions.ts +397 -0
- package/src/commands/ci-setup.ts +381 -0
- package/src/commands/config.ts +99 -0
- package/src/commands/doctor.ts +108 -0
- package/src/commands/init.ts +139 -0
- package/src/commands/mcp.ts +10 -0
- package/src/commands/menu.ts +129 -0
- package/src/commands/post.ts +302 -0
- package/src/commands/skills.ts +83 -0
- package/src/commands/status.ts +173 -0
- package/src/commands/triage.ts +342 -0
- package/src/commands/uninstall.ts +37 -0
- package/src/commands/upgrade.ts +214 -0
- package/src/commands/usage.ts +98 -0
- package/src/compare/compare.ts +28 -0
- package/src/compare/floor-test.ts +471 -0
- package/src/compare/martian-adapter.ts +80 -0
- package/src/config/config.ts +267 -0
- package/src/corpus/corpus.ts +1094 -0
- package/src/corpus/preflight.ts +1789 -0
- package/src/diversity/accounting.ts +90 -0
- package/src/diversity/adjudication.ts +145 -0
- package/src/diversity/admission.ts +60 -0
- package/src/diversity/benchmark.ts +195 -0
- package/src/diversity/clustering.ts +285 -0
- package/src/diversity/errors.ts +7 -0
- package/src/diversity/identity.ts +285 -0
- package/src/diversity/index.ts +11 -0
- package/src/diversity/pipeline-integration.ts +448 -0
- package/src/diversity/projection.ts +57 -0
- package/src/diversity/self-contained.ts +68 -0
- package/src/diversity/target-validation.ts +179 -0
- package/src/doctor-migration.ts +209 -0
- package/src/doctor.ts +723 -0
- package/src/errors.ts +83 -0
- package/src/execution/admission.ts +75 -0
- package/src/execution/atomic-write.ts +35 -0
- package/src/execution/attempt-evidence.ts +415 -0
- package/src/execution/bucket-id.ts +136 -0
- package/src/execution/concurrency-limiter.ts +175 -0
- package/src/execution/contracts.ts +443 -0
- package/src/execution/failure-classification.ts +42 -0
- package/src/execution/failure-policy.ts +240 -0
- package/src/execution/harness.ts +2548 -0
- package/src/execution/settlement.ts +348 -0
- package/src/execution/spawned-process.ts +66 -0
- package/src/execution/spend-limiter.ts +424 -0
- package/src/execution/step-artifacts.ts +55 -0
- package/src/execution/usage-normalized.ts +469 -0
- package/src/git/git.ts +396 -0
- package/src/git/identity.ts +64 -0
- package/src/git/refs.ts +140 -0
- package/src/home-preflight.ts +443 -0
- package/src/home.ts +243 -0
- package/src/ignore-file.ts +336 -0
- package/src/ignore-read.ts +118 -0
- package/src/index.ts +34 -0
- package/src/legacy-launchd.ts +29 -0
- package/src/mcp/mcp.ts +150 -0
- package/src/mcp/preflight.ts +407 -0
- package/src/menu-context.ts +192 -0
- package/src/model/catalog.ts +251 -0
- package/src/model/exec-snapshots.ts +390 -0
- package/src/model/free-discovery.ts +302 -0
- package/src/model/provider-capabilities.ts +814 -0
- package/src/model/routing.ts +600 -0
- package/src/pr/admission.ts +847 -0
- package/src/pr/ci-admission-gate.ts +665 -0
- package/src/pr/ci-publish.ts +76 -0
- package/src/pr/discovery.ts +432 -0
- package/src/pr/inline.ts +507 -0
- package/src/pr/plan.ts +366 -0
- package/src/pr/posting.ts +115 -0
- package/src/pr/pr-status.ts +787 -0
- package/src/pr/pr.ts +3642 -0
- package/src/pr/preflight.ts +1031 -0
- package/src/pr/publish-outcome.ts +361 -0
- package/src/pr/range.ts +119 -0
- package/src/pr/reverts-preflight.ts +606 -0
- package/src/pr/reverts.ts +377 -0
- package/src/pr/review-pr.ts +927 -0
- package/src/pr/status.ts +154 -0
- package/src/pr/target.ts +388 -0
- package/src/pr/teardown.ts +437 -0
- package/src/pr/worktree-setup.ts +60 -0
- package/src/production-runtime.ts +1877 -0
- package/src/progress.ts +487 -0
- package/src/rereview/adjudicate.ts +89 -0
- package/src/rereview/classify.ts +362 -0
- package/src/rereview/identity.ts +211 -0
- package/src/rereview/plan.ts +308 -0
- package/src/rereview/prepare.ts +1025 -0
- package/src/rereview/state.ts +767 -0
- package/src/rereview/verify.ts +241 -0
- package/src/review/boundary.ts +105 -0
- package/src/review/consolidate.ts +246 -0
- package/src/review/dedupe.ts +368 -0
- package/src/review/drafts.ts +448 -0
- package/src/review/findings-conformance.ts +120 -0
- package/src/review/findings.ts +714 -0
- package/src/review/pipeline.ts +3977 -0
- package/src/review/preflight.ts +3575 -0
- package/src/review/prompt-set.ts +177 -0
- package/src/review/proof-refs.ts +187 -0
- package/src/review/report.ts +1821 -0
- package/src/review/review.ts +490 -0
- package/src/review/root-cause.ts +120 -0
- package/src/review/route-preflight.ts +289 -0
- package/src/review/run.ts +601 -0
- package/src/review/scout.ts +323 -0
- package/src/review/size-gate.ts +595 -0
- package/src/review/spec.ts +235 -0
- package/src/review/step-runner.ts +233 -0
- package/src/runner-authority.ts +596 -0
- package/src/security/binding-policy.ts +115 -0
- package/src/security/credential-broker.ts +773 -0
- package/src/security/evidence-redaction.ts +62 -0
- package/src/security/execution-authority.ts +27 -0
- package/src/security/redact.ts +71 -0
- package/src/security/workspace-read-broker.ts +466 -0
- package/src/server/client.ts +190 -0
- package/src/server/preflight.ts +217 -0
- package/src/server/server.ts +327 -0
- package/src/signal-handler.ts +50 -0
- package/src/store/activity.ts +413 -0
- package/src/store/backfill-preflight.ts +49 -0
- package/src/store/backfill.ts +141 -0
- package/src/store/gc-preflight.ts +206 -0
- package/src/store/gc.ts +424 -0
- package/src/store/metrics-preflight.ts +229 -0
- package/src/store/metrics.ts +236 -0
- package/src/store/preflight.ts +748 -0
- package/src/store/store.ts +693 -0
- package/src/system-tools.ts +355 -0
- package/src/transport-registry.ts +1124 -0
- package/src/transports/claude-code-cli.ts +1348 -0
- package/src/transports/opencode-admission.ts +515 -0
- package/src/transports/opencode-client.ts +3328 -0
- package/src/transports/opencode-evidence.ts +557 -0
- package/src/transports/opencode-mcp.ts +248 -0
- package/src/transports/opencode-sdk.ts +2294 -0
- package/src/transports/opencode-server.ts +535 -0
- package/src/triage/adjudication.ts +113 -0
- package/src/triage/reply.ts +125 -0
- package/src/triage/triage.ts +255 -0
- package/src/ui/activity.ts +173 -0
- package/src/ui/config-edit.ts +816 -0
- package/src/ui/config.ts +462 -0
- package/src/ui/menu.ts +1325 -0
- package/src/ui/plan.ts +1039 -0
- package/src/ui/primitives.ts +300 -0
- package/src/ui/progress.ts +298 -0
- package/src/ui/result.ts +552 -0
- package/src/ui/review-menu.ts +312 -0
- package/src/ui/select.ts +528 -0
- package/src/ui/tree.ts +178 -0
- package/src/uninstaller.ts +479 -0
- package/src/updater.ts +836 -0
- package/src/usage.ts +77 -0
- package/src/watch/preflight.ts +1389 -0
- package/src/watch/watch.ts +1235 -0
- package/src/wizard.ts +877 -0
|
@@ -0,0 +1,443 @@
|
|
|
1
|
+
import type { Effort } from "#model/catalog";
|
|
2
|
+
import type {
|
|
3
|
+
CredentialKind,
|
|
4
|
+
ExecutableAllowlistEntry,
|
|
5
|
+
ExecutableDenialCode,
|
|
6
|
+
ProviderCapabilityReport,
|
|
7
|
+
RunnerBackend,
|
|
8
|
+
VerifiedExecutable,
|
|
9
|
+
} from "#model/provider-capabilities";
|
|
10
|
+
import type { StepSpec } from "#review/step-runner";
|
|
11
|
+
import type { CredentialBroker } from "../security/credential-broker";
|
|
12
|
+
import type { WorkspaceDenialCode } from "../security/workspace-read-broker";
|
|
13
|
+
import type {
|
|
14
|
+
NormalizedTokens,
|
|
15
|
+
NormalizedUsage,
|
|
16
|
+
UsageBillingMode,
|
|
17
|
+
} from "./usage-normalized";
|
|
18
|
+
|
|
19
|
+
export type {
|
|
20
|
+
CredentialKind,
|
|
21
|
+
ExecutableAllowlistEntry,
|
|
22
|
+
ExecutableDenialCode,
|
|
23
|
+
ProviderCapabilityReport,
|
|
24
|
+
RunnerBackend,
|
|
25
|
+
VerifiedExecutable,
|
|
26
|
+
WorkspaceDenialCode,
|
|
27
|
+
};
|
|
28
|
+
|
|
29
|
+
export type DenialCode = WorkspaceDenialCode | ExecutableDenialCode;
|
|
30
|
+
|
|
31
|
+
export interface ResolvedModelRoute {
|
|
32
|
+
readonly backend: RunnerBackend;
|
|
33
|
+
readonly provider: string;
|
|
34
|
+
readonly gateway?: string;
|
|
35
|
+
readonly modelFamily: string;
|
|
36
|
+
readonly modelSnapshot: string;
|
|
37
|
+
readonly modelVariant?: string;
|
|
38
|
+
}
|
|
39
|
+
|
|
40
|
+
export interface IsolationProjection {
|
|
41
|
+
readonly credentialProjectionId: string;
|
|
42
|
+
readonly env: Readonly<Record<string, string>>;
|
|
43
|
+
readonly syntheticHome: string;
|
|
44
|
+
readonly syntheticConfigHome: string;
|
|
45
|
+
readonly syntheticTmp: string;
|
|
46
|
+
readonly verifiedBinaryPath: string;
|
|
47
|
+
}
|
|
48
|
+
|
|
49
|
+
export interface TransportRequest {
|
|
50
|
+
readonly sessionId: string;
|
|
51
|
+
readonly attempt: number;
|
|
52
|
+
readonly route: ResolvedModelRoute;
|
|
53
|
+
// The model string the provider CLI/SDK actually receives. For Claude Code
|
|
54
|
+
// direct routes this is the logical alias (sonnet|opus|haiku); configured
|
|
55
|
+
// routes use route.modelSnapshot instead.
|
|
56
|
+
readonly executionModel: string;
|
|
57
|
+
// The effort the engine asked for (StepSpec.effort), always present. Only
|
|
58
|
+
// the Claude Code transport can APPLY it (`--effort`); a backend that cannot
|
|
59
|
+
// still receives it so the evidence can record intent beside the fact that
|
|
60
|
+
// it was not applied.
|
|
61
|
+
readonly effort: Effort;
|
|
62
|
+
readonly systemPromptPath: string;
|
|
63
|
+
readonly systemPromptSha256: string;
|
|
64
|
+
readonly userPrompt: string;
|
|
65
|
+
readonly cwd: string;
|
|
66
|
+
readonly tools: readonly string[];
|
|
67
|
+
readonly mcpConfigPath?: string;
|
|
68
|
+
readonly isolation: IsolationProjection;
|
|
69
|
+
}
|
|
70
|
+
|
|
71
|
+
export interface ProviderTerminalProof {
|
|
72
|
+
readonly eventId: string;
|
|
73
|
+
readonly providerStatus: string;
|
|
74
|
+
readonly providerObservedAt: string;
|
|
75
|
+
readonly exitCode?: number;
|
|
76
|
+
readonly signal?: string;
|
|
77
|
+
}
|
|
78
|
+
|
|
79
|
+
export interface ProviderEventBase {
|
|
80
|
+
readonly sessionId: string;
|
|
81
|
+
readonly attempt: number;
|
|
82
|
+
readonly seq: number;
|
|
83
|
+
readonly observedAt: string;
|
|
84
|
+
}
|
|
85
|
+
|
|
86
|
+
export type ProviderEvent =
|
|
87
|
+
| (ProviderEventBase & { readonly type: "delta"; readonly text: string })
|
|
88
|
+
| (ProviderEventBase & {
|
|
89
|
+
readonly type: "usage";
|
|
90
|
+
readonly mode: "snapshot" | "delta";
|
|
91
|
+
readonly final: boolean;
|
|
92
|
+
// D1-08 PR2 (§8): a streamed usage UPDATE, not a full snapshot — only
|
|
93
|
+
// the leaves this event actually reports are present. cashCostUsd
|
|
94
|
+
// carries a per-event cost delta/snapshot alongside the token leaves;
|
|
95
|
+
// `applyUsageUpdate` (usage-normalized.ts) folds a sequence of these
|
|
96
|
+
// through the same snapshot/delta state machine as the tokens.
|
|
97
|
+
readonly usage: Partial<NormalizedTokens> & {
|
|
98
|
+
readonly cashCostUsd?: number;
|
|
99
|
+
};
|
|
100
|
+
})
|
|
101
|
+
| (ProviderEventBase & {
|
|
102
|
+
readonly type: "diagnostic";
|
|
103
|
+
readonly level: "info" | "warn" | "error";
|
|
104
|
+
readonly message: string;
|
|
105
|
+
})
|
|
106
|
+
| (ProviderEventBase & { readonly type: "heartbeat" })
|
|
107
|
+
| (ProviderEventBase & {
|
|
108
|
+
readonly type: "terminal";
|
|
109
|
+
readonly origin: "provider" | "transport" | "harness";
|
|
110
|
+
readonly proof?: ProviderTerminalProof;
|
|
111
|
+
readonly status: "completed" | "failed" | "cancelled";
|
|
112
|
+
readonly integrity: TransportOutcome["protocolIntegrity"];
|
|
113
|
+
});
|
|
114
|
+
|
|
115
|
+
export interface AsyncEventSink {
|
|
116
|
+
push(event: ProviderEvent): Promise<"accepted" | "closed">;
|
|
117
|
+
close(): void;
|
|
118
|
+
}
|
|
119
|
+
|
|
120
|
+
// #175 half 2, 2026-09-02. WHICH model(s) the provider actually ran, as the
|
|
121
|
+
// provider reported them — not as we asked for them.
|
|
122
|
+
//
|
|
123
|
+
// WHY this exists at all: the engine used to assert a model version up front,
|
|
124
|
+
// from a hardcoded alias -> snapshot mapping that nothing verified (see the
|
|
125
|
+
// #175 note atop model/catalog.ts). The Claude CLI has been reporting the
|
|
126
|
+
// truth in its `modelUsage` block the whole time and we discarded it.
|
|
127
|
+
//
|
|
128
|
+
// WHY it is a LIST and not "the model that ran": verified live 2026-09-02, a
|
|
129
|
+
// single `--model sonnet` invocation reported TWO models — the requested
|
|
130
|
+
// sonnet snapshot AND `claude-haiku-4-5-20251001`, which the CLI uses for its
|
|
131
|
+
// own internal work. Collapsing that to a scalar would be a new false
|
|
132
|
+
// provenance claim replacing the one #175 removed. Nothing may assume the
|
|
133
|
+
// requested model appears here, or that exactly one entry does.
|
|
134
|
+
//
|
|
135
|
+
// `model` is the exact snapshot the provider keyed the entry on;
|
|
136
|
+
// `canonicalModel` is the family it reported for that snapshot, absent when
|
|
137
|
+
// the provider did not say. Neither is derived, defaulted, or cross-checked
|
|
138
|
+
// against the route — a mismatch between what we asked for and what ran is a
|
|
139
|
+
// FACT to record, and reconciling it here would erase it.
|
|
140
|
+
export interface ObservedModel {
|
|
141
|
+
readonly model: string;
|
|
142
|
+
readonly canonicalModel?: string;
|
|
143
|
+
}
|
|
144
|
+
|
|
145
|
+
export interface DiagnosticEvidence {
|
|
146
|
+
readonly schema: string;
|
|
147
|
+
readonly status: "complete" | "incomplete" | "unavailable";
|
|
148
|
+
readonly redactedJson: string;
|
|
149
|
+
}
|
|
150
|
+
|
|
151
|
+
export interface TransportOutcome {
|
|
152
|
+
readonly diagnosticEvidence?: DiagnosticEvidence;
|
|
153
|
+
// A cause the transport holds as a typed fact, not as witness text. resolveFailureCause
|
|
154
|
+
// consults it before the witness classifier, so a fact the stderr witness could only
|
|
155
|
+
// spell keeps its own cause instead of the runtime's generic one.
|
|
156
|
+
readonly failureCause?: TransportFailureCause;
|
|
157
|
+
// One short line saying WHY the attempt failed, in the provider's own words, when the
|
|
158
|
+
// transport holds them as a fact (the Claude CLI's API-error envelope). It travels to
|
|
159
|
+
// the CLI record so the PR can name the real cause; the harness copies it from the final
|
|
160
|
+
// attempt's outcome. Absent on a success and when the transport has nothing to say.
|
|
161
|
+
readonly failureDetail?: string;
|
|
162
|
+
// The secret the failed attempt ran on, when the transport can tell: a CI run names it
|
|
163
|
+
// so the fix says WHICH secret to replace or top up. The per-token variables win over
|
|
164
|
+
// CLAUDE_CODE_OAUTH_TOKEN in the CLI (see credentialOf in the Claude transport).
|
|
165
|
+
readonly failureCredential?: FailureCredential;
|
|
166
|
+
readonly completion: "success" | "failed" | "cancelled";
|
|
167
|
+
readonly protocolIntegrity:
|
|
168
|
+
| "verified"
|
|
169
|
+
| "truncated"
|
|
170
|
+
| "malformed"
|
|
171
|
+
| "overflow"
|
|
172
|
+
| "unverified";
|
|
173
|
+
readonly terminalProof?: ProviderTerminalProof;
|
|
174
|
+
readonly finalText: string;
|
|
175
|
+
// D1-08 PR2 (§8): normalized disjoint usage leaves, not the legacy flat
|
|
176
|
+
// shape. `projectLegacyUsage` (usage-normalized.ts) is the ONLY bridge back
|
|
177
|
+
// to `SessionUsage`, applied at the `runPipeline` return boundary.
|
|
178
|
+
readonly usage: NormalizedUsage;
|
|
179
|
+
readonly stderrTail: string;
|
|
180
|
+
// #126: the transport's OWN diagnostics — counts and tallies that describe
|
|
181
|
+
// how the transport observed the attempt, not what the provider said.
|
|
182
|
+
// `stderrTail` is the classification witness and its declared meaning is
|
|
183
|
+
// "the provider's words"; finalText is already excluded from it for exactly
|
|
184
|
+
// that reason, and our own text is the same hazard from the other
|
|
185
|
+
// direction. Measured, not guessed: a `429` poll-timeout count classified
|
|
186
|
+
// `rate_limit` and a `503` one `network_transient` on the transport's own
|
|
187
|
+
// patterns, while the legacy classifier's unanchored `502|503|529` turned
|
|
188
|
+
// `1502` into a transient too — and the prose "poll round(s) timed out"
|
|
189
|
+
// matched the legacy `timed out` pattern at EVERY count, deterministically.
|
|
190
|
+
// This channel exists so those facts stay visible in the attempt log
|
|
191
|
+
// without any classifier being able to read them. Nothing may classify off
|
|
192
|
+
// it; every witness pattern lives on `stderrTail` alone.
|
|
193
|
+
readonly diagnosticsTail?: string;
|
|
194
|
+
// #214: how many provider tool-call parts this attempt observed.
|
|
195
|
+
// `undefined` — never a fabricated 0 — when the transport cannot observe
|
|
196
|
+
// tools, because 0 is the claim "we looked and the model issued none" and
|
|
197
|
+
// absence is the honest shape for "we were told nothing" (the same rule
|
|
198
|
+
// `observedModels` follows one field down). OpenCode stamps the count
|
|
199
|
+
// after a session actually ran, including 0. Claude Code derives it from
|
|
200
|
+
// the CLI result's `num_turns` (#299), as a lower bound that is exact on
|
|
201
|
+
// zero versus nonzero, and omits it when `num_turns` is not an integer >= 1.
|
|
202
|
+
readonly toolInvocations?: number;
|
|
203
|
+
// #175 half 2: the models the provider says it ran. `undefined` — never
|
|
204
|
+
// `[]` — when the provider reported nothing, because an empty list is the
|
|
205
|
+
// claim "we looked and nothing ran" and absence is the honest shape for
|
|
206
|
+
// "we were told nothing" (the same rule `normalizeUnavailableUsage` follows
|
|
207
|
+
// for tokens one field up).
|
|
208
|
+
//
|
|
209
|
+
// WHERE THIS STOPS, stated here so the next reader does not assume more
|
|
210
|
+
// wiring than exists: the ONLY consumer is the per-attempt log written by
|
|
211
|
+
// `writeAttemptLog` (harness.ts). It is deliberately NOT folded into
|
|
212
|
+
// `NormalizedUsage` — that is §8's numeric contract, summed field-by-field
|
|
213
|
+
// by `sumNormalizedUsage` into a run-level rollup, and a list of names
|
|
214
|
+
// needs union-with-dedup semantics that would have to be designed rather
|
|
215
|
+
// than assumed. It therefore does not reach `StepResult`, `per_agent`,
|
|
216
|
+
// `pipeline.json`, or the terminal report. Wiring it to provenance is a
|
|
217
|
+
// separate slice; parsing it and making it readable is this one.
|
|
218
|
+
readonly observedModels?: readonly ObservedModel[];
|
|
219
|
+
readonly timedOut?: boolean;
|
|
220
|
+
readonly exitCode?: number;
|
|
221
|
+
}
|
|
222
|
+
|
|
223
|
+
// The GitHub secrets a Claude review can run on (see TransportOutcome.failureCredential).
|
|
224
|
+
// The CI record mirrors this list in src/ci/outcome.ts, tied by `satisfies`.
|
|
225
|
+
export type FailureCredential =
|
|
226
|
+
| "ANTHROPIC_API_KEY"
|
|
227
|
+
| "ANTHROPIC_AUTH_TOKEN"
|
|
228
|
+
| "CLAUDE_CODE_OAUTH_TOKEN";
|
|
229
|
+
|
|
230
|
+
export type TransportFailureCause =
|
|
231
|
+
| "auth_invalid"
|
|
232
|
+
| "quota_exhausted"
|
|
233
|
+
| "rate_limit"
|
|
234
|
+
| "network_transient"
|
|
235
|
+
| "protocol_truncation"
|
|
236
|
+
| "protocol_mismatch"
|
|
237
|
+
| "protocol_overflow"
|
|
238
|
+
| "output_limit_exceeded"
|
|
239
|
+
| "context_window_exceeded"
|
|
240
|
+
| "safety_refusal"
|
|
241
|
+
| "provider_configuration_invalid"
|
|
242
|
+
| "model_unavailable"
|
|
243
|
+
| "missing_provider_record"
|
|
244
|
+
| "runtime_unavailable"
|
|
245
|
+
| "remote_abort_unconfirmed";
|
|
246
|
+
|
|
247
|
+
// Every TransportFailureCause as a runtime lookup. The `satisfies` clause fails tsc
|
|
248
|
+
// when a cause joins the union and is missing here, and when a key is not a cause.
|
|
249
|
+
const TRANSPORT_FAILURE_CAUSE_LOOKUP = {
|
|
250
|
+
auth_invalid: true,
|
|
251
|
+
quota_exhausted: true,
|
|
252
|
+
rate_limit: true,
|
|
253
|
+
network_transient: true,
|
|
254
|
+
protocol_truncation: true,
|
|
255
|
+
protocol_mismatch: true,
|
|
256
|
+
protocol_overflow: true,
|
|
257
|
+
output_limit_exceeded: true,
|
|
258
|
+
context_window_exceeded: true,
|
|
259
|
+
safety_refusal: true,
|
|
260
|
+
provider_configuration_invalid: true,
|
|
261
|
+
model_unavailable: true,
|
|
262
|
+
missing_provider_record: true,
|
|
263
|
+
runtime_unavailable: true,
|
|
264
|
+
remote_abort_unconfirmed: true,
|
|
265
|
+
} satisfies Record<TransportFailureCause, true>;
|
|
266
|
+
|
|
267
|
+
export function isTransportFailureCause(
|
|
268
|
+
value: unknown,
|
|
269
|
+
): value is TransportFailureCause {
|
|
270
|
+
return (
|
|
271
|
+
typeof value === "string" &&
|
|
272
|
+
Object.hasOwn(TRANSPORT_FAILURE_CAUSE_LOOKUP, value)
|
|
273
|
+
);
|
|
274
|
+
}
|
|
275
|
+
|
|
276
|
+
// #150 (built on #149's server-lifetime OpenCode credential fix). Declares
|
|
277
|
+
// whether a `ProviderTransport` consumes `StepExecutionHarness.run`'s
|
|
278
|
+
// per-STEP credential projection — the one `credentialBroker.project()`
|
|
279
|
+
// materializes, writes to disk, and destroys once per attempt, and hands
|
|
280
|
+
// the transport as `TransportRequest.isolation` (harness.ts ~728-808).
|
|
281
|
+
//
|
|
282
|
+
// "per-step" (or the field left undeclared) means "yes, this transport
|
|
283
|
+
// reads `request.isolation.env`/`verifiedBinaryPath`, so keep projecting" —
|
|
284
|
+
// today's only behaviour, and the FAIL-SAFE default: a transport nobody
|
|
285
|
+
// updated to declare itself still gets protected credentials rather than
|
|
286
|
+
// silently running on whatever `buildChildEnv` passes through.
|
|
287
|
+
//
|
|
288
|
+
// "server-lifetime" means the transport's OWN process already owns
|
|
289
|
+
// credential protection for its whole run, independent of any one step
|
|
290
|
+
// (the OpenCode SDK transport's server, launched once and brokered by
|
|
291
|
+
// `OpenCodeAuthBroker`/`OpenCodeFreeBroker` per #149) — it never reads
|
|
292
|
+
// `isolation` at all. Declaring it lets the harness skip the per-step
|
|
293
|
+
// projection for that transport entirely: no broker call, no auth file
|
|
294
|
+
// written and destroyed for nothing, and no `missing_subscription_record`
|
|
295
|
+
// degrade warning for a projection that was never attempted (#150's
|
|
296
|
+
// finding: that projection had no consumer and its absence was mislabeled
|
|
297
|
+
// as an "operator-env-fallback", implying a degrade that never happened).
|
|
298
|
+
export type CredentialProjectionConsumption = "per-step" | "server-lifetime";
|
|
299
|
+
|
|
300
|
+
export interface ProviderTransport {
|
|
301
|
+
readonly backend: RunnerBackend;
|
|
302
|
+
readonly billingMode?: UsageBillingMode;
|
|
303
|
+
readonly admissionIdentity?: {
|
|
304
|
+
readonly executable: string;
|
|
305
|
+
readonly provider: string;
|
|
306
|
+
};
|
|
307
|
+
readonly cancellationSemantics?: "process-exit" | "provider-proof";
|
|
308
|
+
// #150: absent = "per-step" (fail-safe default, see
|
|
309
|
+
// `CredentialProjectionConsumption` above). A FUTURE transport that starts
|
|
310
|
+
// reading `isolation.env` must declare "per-step" explicitly or it would
|
|
311
|
+
// silently run unprotected the moment some OTHER transport's declaration
|
|
312
|
+
// taught the harness to skip.
|
|
313
|
+
readonly credentialProjection?: CredentialProjectionConsumption;
|
|
314
|
+
readonly defaultRoute?: ResolvedModelRoute;
|
|
315
|
+
capabilities(): Promise<ProviderCapabilityReport>;
|
|
316
|
+
execute(
|
|
317
|
+
request: TransportRequest,
|
|
318
|
+
context: {
|
|
319
|
+
readonly signal: AbortSignal;
|
|
320
|
+
readonly events: AsyncEventSink;
|
|
321
|
+
},
|
|
322
|
+
): Promise<TransportOutcome>;
|
|
323
|
+
classifyFailure(outcome: TransportOutcome): TransportFailureCause | undefined;
|
|
324
|
+
}
|
|
325
|
+
|
|
326
|
+
export interface StepAdmissionGate {
|
|
327
|
+
admit(spec: StepSpec): Promise<void> | void;
|
|
328
|
+
}
|
|
329
|
+
|
|
330
|
+
export interface AuthEvent {
|
|
331
|
+
readonly kind: "workspace" | "executable";
|
|
332
|
+
readonly status: "approved" | "denied";
|
|
333
|
+
readonly reason?: string;
|
|
334
|
+
}
|
|
335
|
+
|
|
336
|
+
export type BillingMode = "subscription" | "metered";
|
|
337
|
+
|
|
338
|
+
export interface EnvironmentPolicy {
|
|
339
|
+
readonly syntheticHome: boolean;
|
|
340
|
+
readonly workspaceReadBroker: boolean;
|
|
341
|
+
}
|
|
342
|
+
|
|
343
|
+
export interface ToolPolicy {
|
|
344
|
+
readonly allowMapOnly: boolean;
|
|
345
|
+
readonly deniedTools: readonly string[];
|
|
346
|
+
}
|
|
347
|
+
|
|
348
|
+
export interface McpPolicy {
|
|
349
|
+
readonly codegraphOnly: boolean;
|
|
350
|
+
readonly verifiedConfigRequired: boolean;
|
|
351
|
+
}
|
|
352
|
+
|
|
353
|
+
export interface VerifiedExecutableStatus {
|
|
354
|
+
readonly resolved: boolean;
|
|
355
|
+
readonly absolutePath?: string;
|
|
356
|
+
readonly sha256?: string;
|
|
357
|
+
readonly reason?: string;
|
|
358
|
+
}
|
|
359
|
+
|
|
360
|
+
export interface AuthProjectionFacts {
|
|
361
|
+
readonly kind: CredentialKind;
|
|
362
|
+
readonly projectionReady: boolean;
|
|
363
|
+
readonly probe: "passed" | "failed" | "not_run";
|
|
364
|
+
}
|
|
365
|
+
|
|
366
|
+
export interface EnvironmentFacts {
|
|
367
|
+
readonly syntheticHome: boolean;
|
|
368
|
+
readonly enumeratedPassthrough: boolean;
|
|
369
|
+
}
|
|
370
|
+
|
|
371
|
+
export interface IsolationFacts {
|
|
372
|
+
readonly workspaceReadBroker: boolean;
|
|
373
|
+
readonly codegraphPolicy: boolean;
|
|
374
|
+
}
|
|
375
|
+
|
|
376
|
+
export interface ToolMcpFacts {
|
|
377
|
+
readonly allowMapEnforced: boolean;
|
|
378
|
+
readonly mcpIntegrityChecked: boolean;
|
|
379
|
+
}
|
|
380
|
+
|
|
381
|
+
export interface ProtocolFacts {
|
|
382
|
+
readonly terminalProof: boolean;
|
|
383
|
+
readonly boundedEvents: boolean;
|
|
384
|
+
readonly usageMode: "snapshot" | "delta" | "none";
|
|
385
|
+
}
|
|
386
|
+
|
|
387
|
+
export interface UsageFacts {
|
|
388
|
+
readonly normalized: boolean;
|
|
389
|
+
}
|
|
390
|
+
|
|
391
|
+
export interface ExactBindingCapabilityReport {
|
|
392
|
+
readonly routeKey: string;
|
|
393
|
+
readonly backend: RunnerBackend;
|
|
394
|
+
readonly sdk: { readonly available: boolean; readonly version?: string };
|
|
395
|
+
readonly binary: VerifiedExecutableStatus;
|
|
396
|
+
readonly auth: AuthProjectionFacts;
|
|
397
|
+
readonly environment: EnvironmentFacts;
|
|
398
|
+
readonly isolation: IsolationFacts;
|
|
399
|
+
readonly toolsMcp: ToolMcpFacts;
|
|
400
|
+
readonly protocol: ProtocolFacts;
|
|
401
|
+
readonly usage: UsageFacts;
|
|
402
|
+
readonly billing: {
|
|
403
|
+
readonly mode: BillingMode;
|
|
404
|
+
readonly pricingApplicability: "required" | "not_applicable";
|
|
405
|
+
readonly tokenPricingAvailable: boolean;
|
|
406
|
+
readonly cashCostAccountingValid: boolean;
|
|
407
|
+
};
|
|
408
|
+
}
|
|
409
|
+
|
|
410
|
+
export interface RuntimeBindingCredential {
|
|
411
|
+
readonly kind: CredentialKind;
|
|
412
|
+
readonly ref: string;
|
|
413
|
+
readonly broker?: CredentialBroker;
|
|
414
|
+
readonly bucketId: string;
|
|
415
|
+
}
|
|
416
|
+
|
|
417
|
+
export interface RuntimeBinding {
|
|
418
|
+
readonly key: string;
|
|
419
|
+
readonly route: ResolvedModelRoute;
|
|
420
|
+
readonly executable: VerifiedExecutable;
|
|
421
|
+
readonly credential: RuntimeBindingCredential;
|
|
422
|
+
readonly environment: EnvironmentPolicy;
|
|
423
|
+
readonly tools: ToolPolicy;
|
|
424
|
+
readonly mcp: McpPolicy;
|
|
425
|
+
readonly admissionIdentity?: {
|
|
426
|
+
readonly executable: string;
|
|
427
|
+
readonly provider: string;
|
|
428
|
+
};
|
|
429
|
+
readonly cancellationSemantics?: "process-exit" | "provider-proof";
|
|
430
|
+
readonly defaultRoute?: ResolvedModelRoute;
|
|
431
|
+
capabilities(): Promise<ExactBindingCapabilityReport>;
|
|
432
|
+
acquire(
|
|
433
|
+
isolation: IsolationProjection,
|
|
434
|
+
registry: {
|
|
435
|
+
get(backend: RunnerBackend): ProviderTransport;
|
|
436
|
+
},
|
|
437
|
+
): Promise<TransportLease>;
|
|
438
|
+
}
|
|
439
|
+
|
|
440
|
+
export interface TransportLease {
|
|
441
|
+
readonly transport: ProviderTransport;
|
|
442
|
+
dispose(): Promise<void>;
|
|
443
|
+
}
|
|
@@ -0,0 +1,42 @@
|
|
|
1
|
+
// The legacy v1 session-failure classification vocabulary
|
|
2
|
+
// (transient|terminal|format), ported verbatim from deep-review's
|
|
3
|
+
// runner/session.ts.
|
|
4
|
+
//
|
|
5
|
+
// Split out of review/step-runner.ts (architecture guard C2):
|
|
6
|
+
// execution/failure-policy.ts's `resolveFailureCause` falls back to this
|
|
7
|
+
// classifier (step 4: for any outcome a transport's own `classifyFailure`
|
|
8
|
+
// and the parse check both miss), and review/step-runner.ts itself depends
|
|
9
|
+
// on execution/harness.ts for StepExecutionHarness — so failure-policy.ts
|
|
10
|
+
// importing this classifier FROM step-runner.ts closed a value-import
|
|
11
|
+
// cycle back through harness.ts. This file has no imports of its own — a
|
|
12
|
+
// true leaf both sides can depend on one-way.
|
|
13
|
+
|
|
14
|
+
export type FailureClass = "transient" | "terminal" | "format";
|
|
15
|
+
|
|
16
|
+
export function isTransientSessionFailure(result: {
|
|
17
|
+
stderrTail: string;
|
|
18
|
+
resultText: string;
|
|
19
|
+
}): boolean {
|
|
20
|
+
const witness = `${result.stderrTail}\n${result.resultText}`;
|
|
21
|
+
return /API Error|Connection closed|ECONNRESET|socket hang up|timed out|502|503|529|overloaded/i.test(
|
|
22
|
+
witness,
|
|
23
|
+
);
|
|
24
|
+
}
|
|
25
|
+
|
|
26
|
+
export function isTerminalSessionFailure(result: {
|
|
27
|
+
stderrTail: string;
|
|
28
|
+
resultText: string;
|
|
29
|
+
}): boolean {
|
|
30
|
+
const witness = `${result.stderrTail}\n${result.resultText}`;
|
|
31
|
+
return /Not logged in\s*[·.]\s*Please run \/login/i.test(witness);
|
|
32
|
+
}
|
|
33
|
+
|
|
34
|
+
export function classifyFailure(outcome: {
|
|
35
|
+
stderrTail: string;
|
|
36
|
+
resultText: string;
|
|
37
|
+
timedOut: boolean;
|
|
38
|
+
}): FailureClass {
|
|
39
|
+
if (outcome.timedOut) return "transient";
|
|
40
|
+
if (isTerminalSessionFailure(outcome)) return "terminal";
|
|
41
|
+
return isTransientSessionFailure(outcome) ? "transient" : "format";
|
|
42
|
+
}
|