@xzzpig/pi-subagents 0.16.0 → 0.19.0
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/CHANGELOG.md +233 -1
- package/README.md +5 -3
- package/agents/claude-code-writer.md +1 -0
- package/agents/codex-exec-writer.md +1 -0
- package/agents/cursor-agent-writer.md +1 -0
- package/agents/reviewer.md +3 -2
- package/agents/worker.md +3 -2
- package/docs/agents.md +34 -21
- package/docs/configuration.md +82 -7
- package/docs/extension-api.md +75 -8
- package/docs/missions.md +13 -13
- package/docs/models.md +7 -5
- package/docs/observability.md +23 -13
- package/docs/standalone-background.md +5 -5
- package/docs/tool-reference.md +98 -100
- package/docs/watchdog.md +2 -0
- package/docs/workflows.md +214 -162
- package/inspector-runner.mjs +1 -1
- package/node_modules/@xzzpig/pi-components/package.json +4 -4
- package/package.json +15 -9
- package/prompts/council.md +6 -3
- package/prompts/review-loop.md +1 -1
- package/runner-peer-loader.mjs +4 -7
- package/runner-peer-preload.mjs +5 -7
- package/skills/council-mode/SKILL.md +1 -1
- package/skills/pi-subagents/SKILL.md +13 -5
- package/skills/pi-subagents/references/constraints-and-recipes.md +8 -5
- package/skills/pi-subagents/references/execution-controls.md +72 -69
- package/skills/pi-subagents/references/management-authoring-rpc.md +4 -4
- package/skills/pi-subagents/references/multi-lane-orchestration.md +9 -1
- package/skills/pi-subagents/references/prompting-and-roles.md +52 -54
- package/src/agents/advertised-agent-prompt.ts +21 -41
- package/src/agents/agent-eject.ts +3 -3
- package/src/agents/agent-management.ts +63 -54
- package/src/agents/agent-memory.ts +90 -3
- package/src/agents/agent-serializer.ts +4 -4
- package/src/agents/agents.ts +132 -51
- package/src/agents/global-npm-root.ts +58 -0
- package/src/agents/proactive-skills.ts +4 -1
- package/src/agents/runtime-agent-registry.ts +14 -11
- package/src/agents/skills.ts +42 -9
- package/src/api/delegation.ts +15 -0
- package/src/api/inspectors.ts +27 -0
- package/src/api/preflight.ts +7 -2
- package/src/extension/config.ts +12 -4
- package/src/extension/control-notices.ts +3 -0
- package/src/extension/doctor.ts +10 -4
- package/src/extension/fanout-child.ts +20 -3
- package/src/extension/herdr-pi-bridge.ts +2 -0
- package/src/extension/index.ts +168 -206
- package/src/extension/public-execution.ts +69 -35
- package/src/extension/reply-workflow-script.ts +58 -0
- package/src/extension/rpc.ts +39 -11
- package/src/extension/schemas.ts +47 -138
- package/src/extension/subagent-guide.ts +12 -4
- package/src/extension/tool-activation.ts +146 -0
- package/src/extension/tool-description.ts +62 -27
- package/src/extension/upgrade-notice.ts +40 -0
- package/src/inspectors/ghostty/plugin.ts +13 -1
- package/src/inspectors/plugins.ts +70 -3
- package/src/integrations/herdr-status.ts +19 -3
- package/src/intercom/intercom-bridge.ts +2 -0
- package/src/intercom/native-supervisor-channel.ts +26 -16
- package/src/intercom/result-intercom.ts +4 -0
- package/src/runs/background/active-async-capacity.ts +3 -25
- package/src/runs/background/async-execution.ts +60 -95
- package/src/runs/background/async-job-tracker.ts +91 -56
- package/src/runs/background/async-resume.ts +16 -8
- package/src/runs/background/async-retention.ts +46 -16
- package/src/runs/background/async-status.ts +2 -0
- package/src/runs/background/await-async-run.ts +76 -0
- package/src/runs/background/binary-bootstrap.ts +34 -6
- package/src/runs/background/chain-root-attachment.ts +4 -4
- package/src/runs/background/control-channel.ts +18 -2
- package/src/runs/background/fleet-view.ts +51 -7
- package/src/runs/background/notify.ts +98 -7
- package/src/runs/background/pid-namespace.ts +18 -0
- package/src/runs/background/process-terminal-candidate.ts +98 -0
- package/src/runs/background/process-terminal.ts +17 -96
- package/src/runs/background/result-watcher.ts +1 -1
- package/src/runs/background/retained-children.ts +3 -2
- package/src/runs/background/run-child-session.ts +20 -6
- package/src/runs/background/run-id-resolver.ts +10 -5
- package/src/runs/background/run-status.ts +16 -4
- package/src/runs/background/runner-aliases.ts +54 -16
- package/src/runs/background/runner-child-launch.ts +4 -2
- package/src/runs/background/runner-startup-failure.ts +56 -0
- package/src/runs/background/scheduled-runs.ts +5 -5
- package/src/runs/background/stale-run-reconciler.ts +19 -3
- package/src/runs/background/subagent-runner-bootstrap.ts +276 -0
- package/src/runs/background/subagent-runner.ts +111 -288
- package/src/runs/background/wait-tool.ts +2 -0
- package/src/runs/background/workflow-terminal-proof.ts +68 -0
- package/src/runs/foreground/async-stop-action.ts +17 -3
- package/src/runs/foreground/execution.ts +111 -139
- package/src/runs/foreground/prompt-audit.ts +3 -9
- package/src/runs/foreground/subagent-executor.ts +715 -258
- package/src/runs/shared/acceptance.ts +167 -68
- package/src/runs/shared/async-status-projection.ts +5 -0
- package/src/runs/shared/capability-ceiling.ts +4 -1
- package/src/runs/shared/child-hooks.ts +1 -16
- package/src/runs/shared/child-launch.ts +37 -16
- package/src/runs/shared/child-runtime-config.ts +3 -0
- package/src/runs/shared/child-session.ts +144 -8
- package/src/runs/shared/child-tool-plan.ts +38 -116
- package/src/runs/shared/dynamic-fanout.ts +1 -1
- package/src/runs/shared/external-cli-runner.ts +3 -1
- package/src/runs/shared/external-job-bridge.ts +54 -0
- package/src/runs/shared/git-environment.ts +29 -0
- package/src/runs/shared/mcp-config-sources.ts +5 -1
- package/src/runs/shared/mcp-direct-tool-allowlist.ts +104 -12
- package/src/runs/shared/mcp-direct-tool-grant.ts +22 -6
- package/src/runs/shared/model-resolution.ts +35 -10
- package/src/runs/shared/model-scope.ts +28 -9
- package/src/runs/shared/nested-events.ts +12 -0
- package/src/runs/shared/orca-progress-tabs.ts +3 -2
- package/src/runs/shared/parallel-utils.ts +4 -1
- package/src/runs/shared/pi-spawn.ts +1 -3
- package/src/runs/shared/single-output.ts +11 -2
- package/src/runs/shared/structured-output.ts +227 -7
- package/src/runs/shared/subagent-control.ts +2 -16
- package/src/runs/shared/subagent-prompt-runtime.ts +29 -5
- package/src/runs/shared/tool-availability.ts +2 -2
- package/src/runs/shared/tool-budget.ts +21 -0
- package/src/runs/shared/usage-reconciliation.ts +73 -0
- package/src/runs/shared/worktree.ts +36 -9
- package/src/shared/child-session-name.ts +4 -2
- package/src/shared/disabled-features.ts +123 -0
- package/src/shared/extension-context.ts +13 -0
- package/src/shared/fork-context.ts +11 -4
- package/src/shared/launch-contract.ts +1 -1
- package/src/shared/model-info.ts +9 -0
- package/src/shared/pruned-fork.ts +69 -56
- package/src/shared/settings.ts +1 -1
- package/src/shared/types.ts +73 -16
- package/src/slash/delegation-adapters.ts +25 -1
- package/src/slash/prompt-template-bridge.ts +18 -3
- package/src/slash/prompt-workflows.ts +12 -1
- package/src/slash/slash-commands.ts +17 -246
- package/src/slash/subagent-cost.ts +313 -0
- package/src/slash/subagents-admin.ts +4 -5
- package/src/tui/fleet-status.ts +38 -8
- package/src/tui/fleet.ts +21 -6
- package/src/tui/render.ts +166 -71
- package/src/tui/running-tone.ts +28 -0
- package/src/watchdog/diff-tool.ts +47 -12
- package/src/watchdog/permission-arbiter.ts +13 -15
- package/src/watchdog/register-child.ts +13 -5
- package/src/watchdog/register-main.ts +3 -3
- package/src/watchdog/review.ts +16 -16
- package/src/watchdog/runtime.ts +2 -1
- package/src/watchdog/turn-delta.ts +12 -6
- package/src/workflows/chat-progress.ts +2 -2
- package/src/workflows/scripted-workflow.ts +225 -59
- package/src/workflows/structured-workflow-scripts.ts +177 -0
- package/src/workflows/workflow-checklist.ts +6 -0
- package/src/workflows/workflow-child-summary.ts +4 -2
- package/src/workflows/workflow-receipt.ts +8 -2
- package/src/workflows/workflow-resources.ts +37 -12
- package/src/workflows/workflow-reuse.ts +120 -0
- package/src/workflows/workflow-revival.ts +123 -0
- package/src/workflows/workflow-settlement.ts +4 -1
- package/src/runs/shared/completion-evidence.ts +0 -89
- package/src/runs/shared/completion-guard.ts +0 -255
- package/src/runs/shared/llm-intent-arbiter.ts +0 -321
- package/src/runs/shared/task-intent.ts +0 -244
|
@@ -18,6 +18,7 @@ import type {
|
|
|
18
18
|
AcceptanceReviewResult,
|
|
19
19
|
AcceptanceVerifyCommand,
|
|
20
20
|
AcceptanceVerifyResult,
|
|
21
|
+
JsonSchemaObject,
|
|
21
22
|
ResolvedAcceptanceConfig,
|
|
22
23
|
ResolvedAcceptanceGate,
|
|
23
24
|
SingleResult,
|
|
@@ -26,7 +27,7 @@ import type {
|
|
|
26
27
|
} from "../../shared/types.ts";
|
|
27
28
|
import { unresolvedChildWatchdogBlockers } from "../../watchdog/child-status.ts";
|
|
28
29
|
import { isAgentContract } from "./agent-contract.ts";
|
|
29
|
-
import {
|
|
30
|
+
import { validateStructuredOutputValue } from "./structured-output.ts";
|
|
30
31
|
|
|
31
32
|
const LEVEL_RANK: Record<Exclude<AcceptanceLevel, "auto">, number> = {
|
|
32
33
|
none: 0,
|
|
@@ -50,9 +51,9 @@ const VALID_EVIDENCE_KINDS: AcceptanceEvidenceKind[] = [
|
|
|
50
51
|
const VALID_EVIDENCE = new Set<AcceptanceEvidenceKind>(VALID_EVIDENCE_KINDS);
|
|
51
52
|
const ACCEPTANCE_EVIDENCE_HELP = `Supported evidence kinds: ${VALID_EVIDENCE_KINDS.join(", ")}. Example: { level: "checked", evidence: ["commands-run", "changed-files"] }.`;
|
|
52
53
|
const ACCEPTANCE_OBJECT_EXAMPLE = "Example: { level: \"checked\", evidence: [\"commands-run\", \"changed-files\"] }.";
|
|
53
|
-
const ACCEPTANCE_CONFIG_KEYS = new Set(["level", "report", "criteria", "evidence", "verify", "review", "stopRules", "reason"]);
|
|
54
|
+
const ACCEPTANCE_CONFIG_KEYS = new Set(["level", "report", "preserveStagedIndex", "criteria", "evidence", "verify", "review", "stopRules", "reason"]);
|
|
54
55
|
const ACCEPTANCE_GATE_KEYS = new Set(["id", "must", "evidence", "severity"]);
|
|
55
|
-
const ACCEPTANCE_VERIFY_KEYS = new Set(["id", "command", "timeoutMs", "cwd", "env", "allowFailure"]);
|
|
56
|
+
const ACCEPTANCE_VERIFY_KEYS = new Set(["id", "command", "timeoutMs", "cwd", "env", "allowFailure", "output", "schema"]);
|
|
56
57
|
const ACCEPTANCE_REVIEW_KEYS = new Set(["agent", "focus", "required"]);
|
|
57
58
|
const EXPLICIT_REVIEWED_UNAVAILABLE = "is an achieved status, not a requestable acceptance level. For a read-only reviewer call, omit acceptance. To require independent review of a writer result, use acceptance.review.required and orchestrate the reviewer separately.";
|
|
58
59
|
|
|
@@ -86,38 +87,8 @@ function inferLevel(input: {
|
|
|
86
87
|
dynamic?: boolean;
|
|
87
88
|
dynamicGroup?: boolean;
|
|
88
89
|
}): { level: Exclude<AcceptanceLevel, "auto">; reasons: string[]; criteria: string[]; evidence: AcceptanceEvidenceKind[]; review?: { agent?: string; required?: boolean } } {
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
const reasons: string[] = [];
|
|
92
|
-
// Declared roles replace name heuristics, so use the full writer grammar to detect explicit mutation independently of the actual agent name.
|
|
93
|
-
const intent = classifyTaskMutationIntent(input.acceptanceRole ? "worker" : input.agentName, input.task ?? "");
|
|
94
|
-
const readOnlyTask = intent.kind === "read-only"
|
|
95
|
-
|| (intent.kind === "unknown" && /\b(?:read[- ]only|review[- ]only|no edits|without edits|inspect|summari[sz]e)\b/.test(task));
|
|
96
|
-
const rolePatchTask = input.acceptanceRole !== undefined
|
|
97
|
-
&& intent.kind !== "read-only"
|
|
98
|
-
&& !/\b(?:do not|don't|must not)\s+patch\b/.test(task)
|
|
99
|
-
&& /\bpatch\s+(?:(?:\.{0,2}[\\/])?(?:[\w.-]+[\\/])+[\w.-]+|[\w.-]+\.[a-z0-9]+\b|(?:the\s+)?parser\b)/.test(stripSeverityCompounds(task));
|
|
100
|
-
const taskMayWrite = readOnlyTask ? false : taskMayMutate(input.task ?? "") || intent.kind === "implementation" || rolePatchTask;
|
|
101
|
-
const readOnlyAgent = input.acceptanceRole === "read-only"
|
|
102
|
-
|| (input.acceptanceRole === undefined && /\b(?:reviewer|oracle|scout|researcher|analyst)\b/.test(agent));
|
|
103
|
-
const writeTask = taskMayWrite
|
|
104
|
-
|| (input.acceptanceRole === "writer" && !readOnlyTask)
|
|
105
|
-
|| (input.acceptanceRole === undefined && /\bworker\b/.test(agent) && !readOnlyTask);
|
|
106
|
-
const inferredReadOnly = readOnlyTask || ((readOnlyAgent || input.acceptanceRole === "read-only") && !taskMayWrite);
|
|
107
|
-
const roleResolvesReadOnly = input.acceptanceRole !== undefined && inferredReadOnly;
|
|
108
|
-
const dynamicResolvesReadOnly = inferredReadOnly && !writeTask;
|
|
109
|
-
const riskyKeywordPattern = /\b(?:release|migration|migrate|security|data[- ]loss|destructive|post-review|fix pass)\b/;
|
|
110
|
-
// Topic keywords cannot override classified read-only intent; unknown tasks keep their risk gate.
|
|
111
|
-
const keywordRiskReadOnly = input.acceptanceRole === undefined
|
|
112
|
-
? intent.kind === "read-only"
|
|
113
|
-
: inferredReadOnly;
|
|
114
|
-
const risky = Boolean(input.async && writeTask)
|
|
115
|
-
|| (Boolean(input.dynamic) && !roleResolvesReadOnly && !dynamicResolvesReadOnly)
|
|
116
|
-
|| (Boolean(input.dynamicGroup) && !roleResolvesReadOnly && !dynamicResolvesReadOnly)
|
|
117
|
-
|| (!keywordRiskReadOnly && riskyKeywordPattern.test(task));
|
|
118
|
-
|
|
119
|
-
if (risky) {
|
|
120
|
-
reasons.push(input.async ? "async write-capable or risky run" : "risky write-capable run");
|
|
90
|
+
if (input.acceptanceRole === "writer" && (input.async || input.dynamic || input.dynamicGroup)) {
|
|
91
|
+
const reasons = [input.async ? "async declared writer" : "dynamic declared writer"];
|
|
121
92
|
if (input.dynamic || input.dynamicGroup) reasons.push("dynamic fanout context");
|
|
122
93
|
return {
|
|
123
94
|
level: "checked",
|
|
@@ -127,28 +98,25 @@ function inferLevel(input: {
|
|
|
127
98
|
review: { agent: "reviewer", required: true },
|
|
128
99
|
};
|
|
129
100
|
}
|
|
130
|
-
if (
|
|
131
|
-
reasons.push(input.acceptanceRole === "writer" && !taskMayWrite ? "declared writer acceptance role" : "write-capable worker/task");
|
|
101
|
+
if (input.acceptanceRole === "writer") {
|
|
132
102
|
return {
|
|
133
103
|
level: "checked",
|
|
134
|
-
reasons,
|
|
104
|
+
reasons: ["declared writer acceptance role"],
|
|
135
105
|
criteria: ["Implement the requested change without widening scope"],
|
|
136
106
|
evidence: requiredEvidenceForLevel("checked"),
|
|
137
107
|
};
|
|
138
108
|
}
|
|
139
|
-
if (
|
|
140
|
-
reasons.push(input.acceptanceRole === "read-only" && !readOnlyTask ? "declared read-only acceptance role" : readOnlyAgent ? "read-only/reviewer-style agent" : "read-only task wording");
|
|
109
|
+
if (input.acceptanceRole === "read-only") {
|
|
141
110
|
return {
|
|
142
111
|
level: "none",
|
|
143
|
-
reasons,
|
|
112
|
+
reasons: ["declared read-only acceptance role"],
|
|
144
113
|
criteria: ["Return concrete findings with file paths and severity when applicable"],
|
|
145
114
|
evidence: ["review-findings", "residual-risks"],
|
|
146
115
|
};
|
|
147
116
|
}
|
|
148
|
-
reasons.push("default lightweight attestation");
|
|
149
117
|
return {
|
|
150
118
|
level: "attested",
|
|
151
|
-
reasons,
|
|
119
|
+
reasons: ["default lightweight attestation"],
|
|
152
120
|
criteria: ["Return a concise result and residual risks when applicable"],
|
|
153
121
|
evidence: ["manual-notes", "residual-risks"],
|
|
154
122
|
};
|
|
@@ -200,15 +168,59 @@ type GateAcceptanceNormalizationResult =
|
|
|
200
168
|
| { ok: true; acceptance?: AcceptanceInput }
|
|
201
169
|
| { ok: false; error: string };
|
|
202
170
|
|
|
171
|
+
const GATE_OBJECT_KEYS = new Set(["command", "output", "schema", "timeoutMs"]);
|
|
172
|
+
export const GATE_INPUT_ERROR = "gate must be a non-empty command string or { command, output?: \"json\", schema?, timeoutMs? }.";
|
|
173
|
+
|
|
174
|
+
export interface GateObjectInput {
|
|
175
|
+
command: string;
|
|
176
|
+
output?: "json";
|
|
177
|
+
schema?: JsonSchemaObject;
|
|
178
|
+
timeoutMs?: number;
|
|
179
|
+
}
|
|
180
|
+
|
|
181
|
+
/** Validates the two accepted gate shapes without normalizing them. Shared by preflight and the workflow sandbox mirror. */
|
|
182
|
+
export function parseGateInput(gate: unknown): { ok: true; gate: GateObjectInput } | { ok: false; error: string } {
|
|
183
|
+
if (typeof gate === "string") {
|
|
184
|
+
return gate.trim() ? { ok: true, gate: { command: gate.trim() } } : { ok: false, error: GATE_INPUT_ERROR };
|
|
185
|
+
}
|
|
186
|
+
if (!gate || typeof gate !== "object" || Array.isArray(gate)) return { ok: false, error: GATE_INPUT_ERROR };
|
|
187
|
+
const value = gate as Record<string, unknown>;
|
|
188
|
+
for (const key of Object.keys(value)) {
|
|
189
|
+
if (!GATE_OBJECT_KEYS.has(key)) return { ok: false, error: `gate.${key} is not supported.` };
|
|
190
|
+
}
|
|
191
|
+
if (typeof value.command !== "string" || !value.command.trim()) return { ok: false, error: GATE_INPUT_ERROR };
|
|
192
|
+
if (value.output !== undefined && value.output !== "json") return { ok: false, error: "gate.output must be \"json\" when present." };
|
|
193
|
+
if (value.schema !== undefined) {
|
|
194
|
+
if (value.output !== "json") return { ok: false, error: "gate.schema requires gate.output: \"json\"." };
|
|
195
|
+
if (!value.schema || typeof value.schema !== "object" || Array.isArray(value.schema)) return { ok: false, error: "gate.schema must be a JSON Schema object." };
|
|
196
|
+
}
|
|
197
|
+
if (value.timeoutMs !== undefined && (!Number.isInteger(value.timeoutMs) || (value.timeoutMs as number) < 1)) return { ok: false, error: "gate.timeoutMs must be an integer >= 1." };
|
|
198
|
+
const parsed: GateObjectInput = { command: value.command.trim() };
|
|
199
|
+
if (value.output === "json") parsed.output = "json";
|
|
200
|
+
// SAFETY: schema was checked above to be a non-array object; timeoutMs to be an integer >= 1.
|
|
201
|
+
if (value.schema !== undefined) parsed.schema = value.schema as JsonSchemaObject;
|
|
202
|
+
if (value.timeoutMs !== undefined) parsed.timeoutMs = value.timeoutMs as number;
|
|
203
|
+
return { ok: true, gate: parsed };
|
|
204
|
+
}
|
|
205
|
+
|
|
206
|
+
/** True when an acceptance policy declares at least one `output: "json"` verify command, from either the gate shorthand or an explicit verify list. */
|
|
207
|
+
export function acceptanceHasTypedVerify(acceptance: AcceptanceInput | undefined): boolean {
|
|
208
|
+
if (!acceptance || typeof acceptance !== "object") return false;
|
|
209
|
+
return Array.isArray(acceptance.verify) && acceptance.verify.some((command) => command && typeof command === "object" && command.output === "json");
|
|
210
|
+
}
|
|
211
|
+
|
|
212
|
+
export const TYPED_VERIFY_OUTPUT_SCHEMA_CONFLICT = "a typed verify command (output: \"json\") cannot be combined with outputSchema; the child would have two structured-output sources.";
|
|
213
|
+
|
|
203
214
|
export function normalizeGateAcceptance(gate: unknown, acceptance: AcceptanceInput | undefined): GateAcceptanceNormalizationResult {
|
|
204
215
|
if (gate === undefined) {
|
|
205
216
|
if (acceptance === undefined) return { ok: true };
|
|
206
217
|
const normalized = normalizeAcceptanceValue(acceptance);
|
|
207
218
|
return normalized.error ? { ok: false, error: normalized.error } : { ok: true, acceptance: normalized.value as AcceptanceInput };
|
|
208
219
|
}
|
|
209
|
-
|
|
220
|
+
const parsed = parseGateInput(gate);
|
|
221
|
+
if (!parsed.ok) return { ok: false, error: parsed.error };
|
|
210
222
|
if (acceptance !== undefined && acceptance !== false) return { ok: false, error: "gate cannot be combined with acceptance; use one gate command or acceptance.verify." + describeGateAcceptanceConflict(gate, acceptance) };
|
|
211
|
-
return { ok: true, acceptance: { level: "verified", verify: [{ id: "gate",
|
|
223
|
+
return { ok: true, acceptance: { level: "verified", verify: [{ id: "gate", ...parsed.gate }] } };
|
|
212
224
|
}
|
|
213
225
|
|
|
214
226
|
export function describeGateAcceptanceConflict(gate: unknown, acceptance: unknown): string {
|
|
@@ -267,6 +279,12 @@ export function validateAcceptanceInput(input: unknown, pathLabel = "acceptance"
|
|
|
267
279
|
if (value.report !== undefined && value.report !== "on" && value.report !== "off") {
|
|
268
280
|
errors.push(`${pathLabel}.report must be on or off.`);
|
|
269
281
|
}
|
|
282
|
+
if (value.preserveStagedIndex !== undefined && value.preserveStagedIndex !== true) {
|
|
283
|
+
errors.push(`${pathLabel}.preserveStagedIndex must be true when provided.`);
|
|
284
|
+
}
|
|
285
|
+
if (value.preserveStagedIndex === true && value.level !== "checked" && value.level !== "verified") {
|
|
286
|
+
errors.push(`${pathLabel}.preserveStagedIndex requires level checked or verified.`);
|
|
287
|
+
}
|
|
270
288
|
if (value.level === "none" && (typeof value.reason !== "string" || !value.reason.trim())) {
|
|
271
289
|
errors.push(`${pathLabel}.reason is required when level is none.`);
|
|
272
290
|
}
|
|
@@ -348,6 +366,11 @@ export function validateAcceptanceInput(input: unknown, pathLabel = "acceptance"
|
|
|
348
366
|
if (cmd.allowFailure !== undefined && typeof cmd.allowFailure !== "boolean") {
|
|
349
367
|
errors.push(`${pathLabel}.verify[${index}].allowFailure must be a boolean.`);
|
|
350
368
|
}
|
|
369
|
+
if (cmd.output !== undefined && cmd.output !== "json") errors.push(`${pathLabel}.verify[${index}].output must be "json" when present.`);
|
|
370
|
+
if (cmd.schema !== undefined) {
|
|
371
|
+
if (cmd.output !== "json") errors.push(`${pathLabel}.verify[${index}].schema requires output: "json".`);
|
|
372
|
+
if (!cmd.schema || typeof cmd.schema !== "object" || Array.isArray(cmd.schema)) errors.push(`${pathLabel}.verify[${index}].schema must be a JSON Schema object.`);
|
|
373
|
+
}
|
|
351
374
|
}
|
|
352
375
|
}
|
|
353
376
|
if (value.review !== undefined && value.review !== false) {
|
|
@@ -471,6 +494,7 @@ export function resolveEffectiveAcceptance(input: {
|
|
|
471
494
|
inferredReason: [],
|
|
472
495
|
criteria,
|
|
473
496
|
evidence,
|
|
497
|
+
preserveStagedIndex: explicit.preserveStagedIndex,
|
|
474
498
|
verify: explicit.verify ?? [],
|
|
475
499
|
review: explicit.review,
|
|
476
500
|
stopRules: explicit.stopRules ?? [],
|
|
@@ -496,6 +520,7 @@ export function resolveEffectiveAcceptance(input: {
|
|
|
496
520
|
inferredReason: inferred.reasons,
|
|
497
521
|
criteria: level === "none" ? [] : criteria,
|
|
498
522
|
evidence: level === "none" ? [] : evidence,
|
|
523
|
+
preserveStagedIndex: explicit.preserveStagedIndex,
|
|
499
524
|
verify: explicit.verify ?? [],
|
|
500
525
|
review,
|
|
501
526
|
stopRules: explicit.stopRules ?? [],
|
|
@@ -527,6 +552,9 @@ export function formatAcceptancePrompt(acceptance: ResolvedAcceptanceConfig, opt
|
|
|
527
552
|
"",
|
|
528
553
|
`Required evidence: ${acceptance.evidence.join(", ") || "none"}`,
|
|
529
554
|
];
|
|
555
|
+
if (acceptance.preserveStagedIndex) {
|
|
556
|
+
lines.push("The host will verify that the staged index is unchanged from launch; report noStagedFiles truthfully even when the preserved index is non-empty.");
|
|
557
|
+
}
|
|
530
558
|
if (acceptance.verify.length > 0) {
|
|
531
559
|
lines.push("", "Runtime verification commands configured by parent:");
|
|
532
560
|
for (const command of acceptance.verify) lines.push(`- ${command.id}: ${command.command}`);
|
|
@@ -1079,9 +1107,32 @@ function checkNoStagedFiles(cwd: string): AcceptanceRuntimeCheck {
|
|
|
1079
1107
|
: { id: "no-staged-files", status: "failed", message: `Staged files present: ${staged.join(", ")}` };
|
|
1080
1108
|
}
|
|
1081
1109
|
|
|
1110
|
+
/** Capture the repository-wide index tree. Throws when Git cannot provide trustworthy evidence. */
|
|
1111
|
+
export function captureStagedIndexBaseline(cwd: string): string {
|
|
1112
|
+
const result = spawnSync("git", ["write-tree"], { cwd, encoding: "utf-8", windowsHide: true });
|
|
1113
|
+
const oid = result.stdout.trim();
|
|
1114
|
+
if (result.status !== 0 || !oid) {
|
|
1115
|
+
const detail = result.stderr.trim();
|
|
1116
|
+
throw new Error(`Unable to capture staged index baseline${detail ? `: ${detail}` : "."}`);
|
|
1117
|
+
}
|
|
1118
|
+
return oid;
|
|
1119
|
+
}
|
|
1120
|
+
|
|
1121
|
+
function checkStagedIndexUnchanged(cwd: string, baseline: string): AcceptanceRuntimeCheck {
|
|
1122
|
+
try {
|
|
1123
|
+
const terminal = captureStagedIndexBaseline(cwd);
|
|
1124
|
+
return terminal === baseline
|
|
1125
|
+
? { id: "staged-index-unchanged", status: "passed", message: "Staged index matches the launch baseline." }
|
|
1126
|
+
: { id: "staged-index-unchanged", status: "failed", message: "Staged index changed after launch." };
|
|
1127
|
+
} catch (error) {
|
|
1128
|
+
return { id: "staged-index-unchanged", status: "failed", message: error instanceof Error ? error.message : String(error) };
|
|
1129
|
+
}
|
|
1130
|
+
}
|
|
1131
|
+
|
|
1082
1132
|
function runStructuralChecks(acceptance: ResolvedAcceptanceConfig, report: AcceptanceReport, cwd: string): AcceptanceRuntimeCheck[] {
|
|
1083
1133
|
const checks: AcceptanceRuntimeCheck[] = [];
|
|
1084
1134
|
for (const kind of acceptance.evidence) {
|
|
1135
|
+
if (kind === "no-staged-files" && acceptance.preserveStagedIndex) continue;
|
|
1085
1136
|
if (kind === "no-staged-files" && report.noStagedFiles === undefined) continue;
|
|
1086
1137
|
const status = reportEvidenceStatus(report, kind);
|
|
1087
1138
|
checks.push({
|
|
@@ -1094,7 +1145,7 @@ function runStructuralChecks(acceptance: ResolvedAcceptanceConfig, report: Accep
|
|
|
1094
1145
|
: `${kind} evidence missing from child report.`,
|
|
1095
1146
|
});
|
|
1096
1147
|
}
|
|
1097
|
-
if (acceptance.evidence.includes("no-staged-files")) checks.push(checkNoStagedFiles(cwd));
|
|
1148
|
+
if (!acceptance.preserveStagedIndex && acceptance.evidence.includes("no-staged-files")) checks.push(checkNoStagedFiles(cwd));
|
|
1098
1149
|
return checks;
|
|
1099
1150
|
}
|
|
1100
1151
|
|
|
@@ -1202,6 +1253,43 @@ function isCachedVerifyResult(value: unknown): value is AcceptanceVerifyResult {
|
|
|
1202
1253
|
&& typeof result.durationMs === "number";
|
|
1203
1254
|
}
|
|
1204
1255
|
|
|
1256
|
+
const TYPED_VERIFY_OUTPUT_MAX_BYTES = 12_000;
|
|
1257
|
+
|
|
1258
|
+
/**
|
|
1259
|
+
* A passing `output: "json"` command must print one JSON document on stdout.
|
|
1260
|
+
* The parsed value becomes the run's structured output; anything else turns
|
|
1261
|
+
* the passing run into a failed one so acceptance rejects rather than
|
|
1262
|
+
* silently dropping the verdict. Stdout is already bounded by trimOutput, so a
|
|
1263
|
+
* truncated document is detected by its marker instead of a parse guess.
|
|
1264
|
+
*/
|
|
1265
|
+
async function applyTypedVerifyOutput(command: AcceptanceVerifyCommand, run: AcceptanceVerifyResult): Promise<void> {
|
|
1266
|
+
if (run.status !== "passed") return;
|
|
1267
|
+
const stdout = run.stdout ?? "";
|
|
1268
|
+
const fail = (message: string): void => {
|
|
1269
|
+
run.status = "failed";
|
|
1270
|
+
run.structuredOutputError = message;
|
|
1271
|
+
};
|
|
1272
|
+
if (!stdout.trim()) return fail("output: \"json\" command printed nothing on stdout.");
|
|
1273
|
+
if (stdout.endsWith("...[truncated]")) return fail(`output: "json" stdout exceeded ${TYPED_VERIFY_OUTPUT_MAX_BYTES} characters and was truncated.`);
|
|
1274
|
+
let value: unknown;
|
|
1275
|
+
try {
|
|
1276
|
+
value = JSON.parse(stdout);
|
|
1277
|
+
} catch (error) {
|
|
1278
|
+
return fail(`output: "json" stdout is not valid JSON: ${error instanceof Error ? error.message : String(error)}`);
|
|
1279
|
+
}
|
|
1280
|
+
if (command.schema) {
|
|
1281
|
+
const validation = await validateStructuredOutputValue(command.schema, value);
|
|
1282
|
+
if (validation.status === "invalid") return fail(`output: "json" stdout does not match gate.schema: ${validation.message}`);
|
|
1283
|
+
}
|
|
1284
|
+
run.structuredOutput = value;
|
|
1285
|
+
}
|
|
1286
|
+
|
|
1287
|
+
/** The first typed verify result on a ledger, if any command produced one. */
|
|
1288
|
+
export function typedVerifyOutput(ledger: AcceptanceLedger | undefined): { value: unknown } | undefined {
|
|
1289
|
+
const run = ledger?.verifyRuns?.find((entry) => entry.structuredOutput !== undefined);
|
|
1290
|
+
return run ? { value: run.structuredOutput } : undefined;
|
|
1291
|
+
}
|
|
1292
|
+
|
|
1205
1293
|
async function runMemoizedVerifyCommand(command: AcceptanceVerifyCommand, defaultCwd: string, options: {
|
|
1206
1294
|
signal?: AbortSignal;
|
|
1207
1295
|
abortMessage?: string;
|
|
@@ -1215,7 +1303,9 @@ async function runMemoizedVerifyCommand(command: AcceptanceVerifyCommand, defaul
|
|
|
1215
1303
|
} catch {
|
|
1216
1304
|
workspaceState = undefined;
|
|
1217
1305
|
}
|
|
1218
|
-
|
|
1306
|
+
// Typed gates read inputs (a report file, a log) that can change without
|
|
1307
|
+
// the tracked tree changing, so their verdicts are never reused.
|
|
1308
|
+
if (!workspaceState || !options.artifactsDir || !options.runId || command.output === "json") {
|
|
1219
1309
|
return runVerifyCommand(command, defaultCwd, options);
|
|
1220
1310
|
}
|
|
1221
1311
|
const envKeys = Object.keys(command.env ?? {}).sort();
|
|
@@ -1387,6 +1477,8 @@ export async function evaluateAcceptance(input: {
|
|
|
1387
1477
|
acceptance: ResolvedAcceptanceConfig;
|
|
1388
1478
|
output: string;
|
|
1389
1479
|
cwd: string;
|
|
1480
|
+
/** Host-captured launch index tree; required by preserveStagedIndex. */
|
|
1481
|
+
stagedIndexBaseline?: string;
|
|
1390
1482
|
/**
|
|
1391
1483
|
* Content the child sent to its configured output file (from its own write
|
|
1392
1484
|
* tool calls, not from disk, so a concurrent writer to the same path cannot
|
|
@@ -1482,30 +1574,36 @@ export async function evaluateAcceptance(input: {
|
|
|
1482
1574
|
ledger.runtimeChecks.push({ id: "verification-config", status: "failed", message: "verified acceptance requires runtime verify commands." });
|
|
1483
1575
|
ledger.status = "rejected";
|
|
1484
1576
|
ledger.evidenceStatus = "rejected";
|
|
1485
|
-
|
|
1486
|
-
|
|
1487
|
-
|
|
1488
|
-
|
|
1489
|
-
|
|
1490
|
-
|
|
1491
|
-
|
|
1492
|
-
|
|
1493
|
-
|
|
1494
|
-
|
|
1495
|
-
|
|
1496
|
-
|
|
1497
|
-
|
|
1498
|
-
ledger.status
|
|
1499
|
-
|
|
1500
|
-
|
|
1501
|
-
|
|
1502
|
-
|
|
1503
|
-
|
|
1504
|
-
|
|
1577
|
+
} else {
|
|
1578
|
+
ledger.verifyRuns = [];
|
|
1579
|
+
for (const command of acceptance.verify) {
|
|
1580
|
+
const run = await runMemoizedVerifyCommand(command, input.cwd, {
|
|
1581
|
+
signal: input.signal,
|
|
1582
|
+
abortMessage: input.abortMessage,
|
|
1583
|
+
artifactsDir: input.artifactsDir,
|
|
1584
|
+
runId: input.runId,
|
|
1585
|
+
});
|
|
1586
|
+
if (command.output === "json") await applyTypedVerifyOutput(command, run);
|
|
1587
|
+
ledger.verifyRuns.push(run);
|
|
1588
|
+
if (input.signal?.aborted) break;
|
|
1589
|
+
}
|
|
1590
|
+
if (ledger.verifyRuns.some((run) => run.status === "failed" || run.status === "timed-out")) {
|
|
1591
|
+
ledger.status = "rejected";
|
|
1592
|
+
ledger.evidenceStatus = "rejected";
|
|
1593
|
+
} else if (!ledger.runtimeChecks.some((check) => check.status === "failed")) {
|
|
1594
|
+
ledger.status = "verified";
|
|
1595
|
+
ledger.evidenceStatus = "verified";
|
|
1596
|
+
}
|
|
1505
1597
|
}
|
|
1506
1598
|
}
|
|
1507
1599
|
|
|
1508
|
-
if (
|
|
1600
|
+
if (acceptance.preserveStagedIndex) {
|
|
1601
|
+
ledger.runtimeChecks.push(input.stagedIndexBaseline
|
|
1602
|
+
? checkStagedIndexUnchanged(input.cwd, input.stagedIndexBaseline)
|
|
1603
|
+
: { id: "staged-index-unchanged", status: "failed", message: "Staged index launch baseline is unavailable." });
|
|
1604
|
+
}
|
|
1605
|
+
|
|
1606
|
+
if (ledger.status === "rejected" || ledger.runtimeChecks.some((check) => check.status === "failed")) {
|
|
1509
1607
|
ledger.status = "rejected";
|
|
1510
1608
|
ledger.evidenceStatus = "rejected";
|
|
1511
1609
|
return ledger;
|
|
@@ -1559,6 +1657,7 @@ export function acceptanceFailureMessage(ledger: AcceptanceLedger): string | und
|
|
|
1559
1657
|
const failedCheck = ledger.runtimeChecks.find((check) => check.status === "failed");
|
|
1560
1658
|
if (failedCheck) return `Acceptance rejected: ${failedCheck.message}`;
|
|
1561
1659
|
const failedVerify = ledger.verifyRuns.find((run) => run.status === "failed" || run.status === "timed-out");
|
|
1660
|
+
if (failedVerify?.structuredOutputError) return `Acceptance verification '${failedVerify.id}' failed: ${failedVerify.structuredOutputError}`;
|
|
1562
1661
|
if (failedVerify) return `Acceptance verification '${failedVerify.id}' ${failedVerify.status}.`;
|
|
1563
1662
|
if (ledger.reviewResult?.status === "blockers") return "Acceptance review found blockers.";
|
|
1564
1663
|
return "Acceptance rejected.";
|
|
@@ -1,5 +1,6 @@
|
|
|
1
1
|
import { sanitizeDisplayText, truncateDisplayText } from "../../shared/display-text.ts";
|
|
2
2
|
import { formatModelThinking } from "../../shared/formatters.ts";
|
|
3
|
+
import { childThinkingLevel, type ThinkingLevel } from "../../shared/model-info.ts";
|
|
3
4
|
import type { AsyncJobState, AsyncJobStep, HostStepFreshness, HostStepMonitorKind, HostStepNode, HostStepState, HostStepVerdict, NestedRunSummary, NestedStepSummary, SubagentRunMode, WorkflowGraphSnapshot, WorkflowPreflightLane, WorkflowPreflight } from "../../shared/types.ts";
|
|
4
5
|
import { HOST_STEP_MAX_COUNT, HOST_STEP_MAX_DETAIL_CHARS, HOST_STEP_MAX_LABEL_CHARS, HOST_STEP_MAX_PROVIDER_CHARS, HOST_STEP_MAX_REASON_CHARS, HOST_STEP_MAX_REF_CHARS, HOST_STEP_MAX_ROLE_CHARS, HOST_STEP_MAX_TARGET_CHARS, hostStepReportName, parseHostStepNode, validHostStepNodes } from "./host-step-status.ts";
|
|
5
6
|
import { workflowPreflightLaneForRuntimeKey } from "../../workflows/workflow-preflight.ts";
|
|
@@ -106,6 +107,8 @@ export interface AsyncStatusWorkflowRow {
|
|
|
106
107
|
kind?: HostStepMonitorKind;
|
|
107
108
|
context?: AsyncJobStep["context"];
|
|
108
109
|
modelThinking?: string;
|
|
110
|
+
/** Thinking level of the child a child row stands for. */
|
|
111
|
+
thinking?: ThinkingLevel;
|
|
109
112
|
activity?: string;
|
|
110
113
|
startedAt?: number;
|
|
111
114
|
endedAt?: number;
|
|
@@ -623,12 +626,14 @@ export function projectAsyncWorkflowRows(
|
|
|
623
626
|
|
|
624
627
|
function projectLoadedWorkflowRow(step: AsyncJobStep, index: number, preflight?: WorkflowPreflightLane): AsyncStatusWorkflowRow {
|
|
625
628
|
const modelThinking = formatModelThinking(step.model, step.thinking) || undefined;
|
|
629
|
+
const thinking = childThinkingLevel(step);
|
|
626
630
|
const activity = workflowStepActivity(step);
|
|
627
631
|
return {
|
|
628
632
|
name: workflowStepName(step, index),
|
|
629
633
|
state: step.status,
|
|
630
634
|
...(step.context ? { context: step.context } : {}),
|
|
631
635
|
...(modelThinking ? { modelThinking } : {}),
|
|
636
|
+
...(thinking ? { thinking } : {}),
|
|
632
637
|
...(activity ? { activity } : {}),
|
|
633
638
|
...(step.startedAt !== undefined ? { startedAt: step.startedAt } : {}),
|
|
634
639
|
...(step.endedAt !== undefined ? { endedAt: step.endedAt } : {}),
|
|
@@ -30,7 +30,6 @@ export interface SubagentCapabilityAudit {
|
|
|
30
30
|
agentAllowed: boolean;
|
|
31
31
|
agentRestrictionSources?: string[];
|
|
32
32
|
/** Builtin tools declared but unavailable on the host runtime. */
|
|
33
|
-
unavailableHostBuiltins?: string[];
|
|
34
33
|
}
|
|
35
34
|
|
|
36
35
|
export interface RegisterSubagentCapabilityCeilingOptions {
|
|
@@ -64,6 +63,10 @@ function validateText(value: unknown, field: string): string {
|
|
|
64
63
|
return value.trim();
|
|
65
64
|
}
|
|
66
65
|
|
|
66
|
+
export function normalizeCapabilityCeilingAllowedAgents(values: unknown): string[] {
|
|
67
|
+
return normalizeCeiling({ allowedAgents: values } as SubagentCapabilityCeiling).allowedAgents!;
|
|
68
|
+
}
|
|
69
|
+
|
|
67
70
|
function normalizeCeiling(ceiling: SubagentCapabilityCeiling): ResolvedSubagentCapabilityCeiling {
|
|
68
71
|
if (!ceiling || typeof ceiling !== "object" || Array.isArray(ceiling)) throw new Error("Invalid capability ceiling; expected an object.");
|
|
69
72
|
const hasAllowedTools = Object.hasOwn(ceiling, "allowedTools");
|
|
@@ -7,7 +7,6 @@ import registerSubagentPromptRuntime from "./subagent-prompt-runtime.ts";
|
|
|
7
7
|
import type { ChildRuntimeConfig } from "./child-runtime-config.ts";
|
|
8
8
|
import type { ChildToolDiagnostic } from "./tool-availability.ts";
|
|
9
9
|
import type { ChildSessionLaunch } from "./child-session.ts";
|
|
10
|
-
import type { ArbiterModelContext } from "./llm-intent-arbiter.ts";
|
|
11
10
|
import type { ChildTranscriptWriter } from "../../shared/child-transcript.ts";
|
|
12
11
|
import { projectRuntimeAcknowledgedExtensions } from "./runtime-acknowledged-extensions.ts";
|
|
13
12
|
|
|
@@ -128,10 +127,9 @@ export function createChildHooks(config: ChildRuntimeConfig): ChildHookExtension
|
|
|
128
127
|
}
|
|
129
128
|
|
|
130
129
|
/** Launch-owned bookkeeping, paired with the same private hook certificate (no callback registration API). */
|
|
131
|
-
export function createCapturedChildHooks(config: ChildRuntimeConfig
|
|
130
|
+
export function createCapturedChildHooks(config: ChildRuntimeConfig) {
|
|
132
131
|
let diagnostic: ChildToolDiagnostic | undefined;
|
|
133
132
|
let acknowledgedIds: string[] | undefined;
|
|
134
|
-
let completionIntentContext: ArbiterModelContext | undefined;
|
|
135
133
|
let finalDrainHeld = false;
|
|
136
134
|
const capture: OwnedCapture = {
|
|
137
135
|
toolDiagnostic: (value) => { diagnostic = value; },
|
|
@@ -139,21 +137,8 @@ export function createCapturedChildHooks(config: ChildRuntimeConfig, runner = fa
|
|
|
139
137
|
};
|
|
140
138
|
Object.assign(config, capture);
|
|
141
139
|
const hooks = childHooks(config, capture, (held) => { finalDrainHeld = held; });
|
|
142
|
-
if (runner) {
|
|
143
|
-
hooks.push({ name: "pi-subagents:completion-intent", factory: (pi) => pi.on("session_start", (_event, childCtx) => {
|
|
144
|
-
// Retain only attempt model services and the session id string, not the live child session.
|
|
145
|
-
completionIntentContext = {
|
|
146
|
-
model: childCtx.model,
|
|
147
|
-
modelRegistry: childCtx.modelRegistry,
|
|
148
|
-
sessionId: childCtx.sessionManager.getSessionId(),
|
|
149
|
-
};
|
|
150
|
-
}) });
|
|
151
|
-
const proof = promptProofs.get(hooks[0]!.factory);
|
|
152
|
-
if (proof) proof.factories = hooks.map((hook) => hook.factory);
|
|
153
|
-
}
|
|
154
140
|
return {
|
|
155
141
|
hooks,
|
|
156
|
-
completionIntentContext: () => completionIntentContext,
|
|
157
142
|
toolDiagnostic: () => diagnostic,
|
|
158
143
|
runtimeAcknowledgedExtensions: () => acknowledgedIds ? projectRuntimeAcknowledgedExtensions(acknowledgedIds) : undefined,
|
|
159
144
|
finalDrainHeld: () => finalDrainHeld,
|