Press n or j to go to the next uncovered block, b, p or k for the previous block.
| 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 | 11x 10x 7x 1x 5x 5x 5x 2x 3x 1x 2x 1x 1x 33x 1x 1x 5x 5x 1x 1x 1x 1x 14x 14x 14x 14x 14x 14x 14x 14x 14x 14x 14x 14x 11x 11x 3x 2x 3x 11x 11x 1x 1x 1x 1x 1x 10x 10x 10x 14x 14x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 14x 14x 20x 14x 14x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 10x 14x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 5x 14x 5x 1x 5x 5x 1x 5x 5x 5x 5x 5x 5x 5x 5x 3x 3x 3x 3x 3x 3x 3x 3x 18x 94x 94x 94x 6x 3x 2x 2x 17x 8x 8x 9x 3x 2x 17x 17x 17x 17x 4x 2x 12x 9x 8x 8x 8x 8x 8x 8x 8x 8x 3x 4x 4x 1x 8x 8x 1x 2x 2x 2x 2x 2x 2x 2x 2x 2x 2x 2x 1x 1x 1x 1x 1x 1x 1x 2x 2x 1x 1x 3x 3x 3x 3x 1x 1x 1x 1x 1x 1x 6x 6x 10x 4x 4x 6x 2x 4x 6x 4x 2x 6x 4x 4x 4x 4x 4x 2x 2x 2x 1x 4x 4x 4x 4x 4x 4x 4x 4x | /**
* Task execution for skills.
* Callback-based state updates for CLI and Ink rendering.
*
* Reporter spec: specs/reporters.md
*/
import type { SkillReport, SeverityThreshold, ConfidenceThreshold, Finding, UsageStats, EventContext, HunkFailure, AuxiliaryUsageMap, ErrorCode } from '../../types/index.js';
import type { SkillDefinition } from '../../config/schema.js';
import { Sentry, emitSkillMetrics, logger } from '../../sentry.js';
import { SkillRunnerError, WardenAuthenticationError, classifyError } from '../../sdk/errors.js';
import {
prepareFiles,
analyzeFile,
aggregateUsage,
aggregateAuxiliaryUsage,
postProcessFindings,
generateSummary,
type AuxiliaryUsageEntry,
type SkillRunnerOptions,
type FileAnalysisCallbacks,
type PreparedFile,
type PRPromptContext,
type ChunkAnalysisResult,
type FindingProcessingEvent,
} from '../../sdk/runner.js';
import { ProviderFailureCircuitBreaker } from '../../sdk/circuit-breaker.js';
import { buildFileReports } from '../../sdk/report-files.js';
import chalk from 'chalk';
import figures from 'figures';
import { Verbosity } from './verbosity.js';
import type { OutputMode } from './tty.js';
import { ICON_CHECK, ICON_SKIPPED } from './icons.js';
import { timestamp } from './tty.js';
import { formatDuration, formatCost, formatLocation, formatSeverityPlain, formatFindingCountsPlain, countBySeverity, pluralize } from './formatters.js';
import { runPool, Semaphore } from '../../utils/index.js';
/**
* Result from processing a single file within a skill task.
*/
interface FileProcessResult {
findings: Finding[];
usage?: UsageStats;
durationMs: number;
failedHunks: number;
failedExtractions: number;
hunkFailures: HunkFailure[];
auxiliaryUsage?: AuxiliaryUsageEntry[];
}
function allAnalysisFailuresHaveCode(
hunkFailures: HunkFailure[],
code: ErrorCode,
): boolean {
const analysisFailures = hunkFailures.filter((failure) => failure.type === 'analysis');
return (
analysisFailures.length > 0
&& analysisFailures.every((failure) => failure.code === code)
);
}
function firstAnalysisFailureMessage(hunkFailures: HunkFailure[], code: ErrorCode): string | undefined {
return hunkFailures.find((failure) => failure.type === 'analysis' && failure.code === code)?.message;
}
function summarizeRunFailure(args: {
totalHunks: number;
hunkFailures: HunkFailure[];
circuitReason?: { code: ErrorCode; message: string };
runtime?: SkillRunnerOptions['runtime'];
}): { code: ErrorCode; message: string } {
const { totalHunks, hunkFailures, circuitReason, runtime } = args;
Iif (circuitReason) {
return circuitReason;
}
if (allAnalysisFailuresHaveCode(hunkFailures, 'auth_failed')) {
return {
code: 'auth_failed',
message: 'Authentication failed. Warden stopped early.',
};
}
if (allAnalysisFailuresHaveCode(hunkFailures, 'invalid_model_selector')) {
return {
code: 'invalid_model_selector',
message: firstAnalysisFailureMessage(hunkFailures, 'invalid_model_selector') ?? 'Invalid Pi model selector.',
};
}
if (allAnalysisFailuresHaveCode(hunkFailures, 'provider_unavailable')) {
return {
code: 'provider_unavailable',
message: `Provider unavailable: all ${totalHunks} chunk${totalHunks === 1 ? '' : 's'} failed to analyze. Warden stopped early.`,
};
}
return {
code: 'all_hunks_failed',
message:
`All ${totalHunks} chunk${totalHunks === 1 ? '' : 's'} failed to analyze. ` +
`This usually indicates an authentication problem. ` +
((runtime ?? 'pi') === 'claude'
? `Verify WARDEN_ANTHROPIC_API_KEY is set correctly, or run 'claude login' when using the Claude runtime without an API key.`
: `Verify WARDEN_MODEL and the WARDEN-prefixed provider API key for that model are set correctly.`),
};
}
/**
* Write a log-mode message to stderr with timestamp prefix.
* Used for non-TTY / plain output.
*/
function logPlain(message: string): void {
console.error(`[${timestamp()}] warden: ${message}`);
}
/**
* Write a debug-level message to stderr.
* Uses chalk.dim formatting in TTY mode, timestamped "DEBUG:" prefix otherwise.
*/
function debugLog(mode: OutputMode, message: string): void {
Iif (mode.isTTY) {
console.error(chalk.dim(`[debug] ${message}`));
} else {
logPlain(`DEBUG: ${message}`);
}
}
/**
* Format a finding's location as a compact string, falling back to 'unknown'.
*/
function findingLocation(finding: Finding): string {
Iif (!finding.location) return 'unknown';
return formatLocation(finding.location.path, finding.location.startLine, finding.location.endLine);
}
function findingSummary(finding: Finding): string {
return `${findingLocation(finding)}: ${finding.title}`;
}
function formatFindingProcessingEvent(event: FindingProcessingEvent): string {
const reason = event.reason ? ` (${event.reason})` : '';
const replacement = event.replacement ? ` -> ${findingSummary(event.replacement)}` : '';
return `${event.stage}:${event.action} ${findingSummary(event.finding)}${replacement}${reason}`;
}
/**
* State of a file being processed by a skill.
*/
export interface FileState {
filename: string;
status: 'pending' | 'running' | 'done' | 'skipped';
currentHunk: number;
totalHunks: number;
findings: Finding[];
usage?: UsageStats;
durationMs?: number;
}
/**
* State of a skill being executed.
*/
export interface SkillState {
name: string;
displayName: string;
status: 'pending' | 'running' | 'done' | 'skipped' | 'error';
startTime?: number;
durationMs?: number;
files: FileState[];
findings: Finding[];
usage?: UsageStats;
auxiliaryUsage?: AuxiliaryUsageMap;
error?: string;
}
/**
* Result from running a skill task.
*/
export interface SkillTaskResult {
name: string;
report?: SkillReport;
failOn?: SeverityThreshold;
minConfidence?: ConfidenceThreshold;
error?: unknown;
}
/**
* Options for creating a skill task.
*/
export interface SkillTaskOptions {
name: string;
displayName?: string;
triggerName?: string;
failOn?: SeverityThreshold;
minConfidence?: ConfidenceThreshold;
/** Resolve the skill definition (may be async for loading) */
resolveSkill: () => Promise<SkillDefinition>;
/** The event context with files to analyze */
context: EventContext;
/** Options passed to the runner */
runnerOptions?: SkillRunnerOptions;
}
/**
* Options for running skill tasks.
*/
export interface RunTasksOptions {
mode: OutputMode;
verbosity: Verbosity;
concurrency: number;
/** Controller that fires when fail-fast detects a finding. Created by caller. */
failFastController?: AbortController;
/** Hook fired after each skill finishes; used by the CLI to stream JSONL to disk. */
onSkillComplete?: (report: SkillReport) => void;
/** Hook fired after each chunk finishes; used by the CLI to stream JSONL to disk. */
onChunkComplete?: (skillName: string, chunk: ChunkAnalysisResult) => void;
}
/**
* Callbacks for reporting skill execution progress to the UI.
*/
export interface SkillProgressCallbacks {
onSkillStart: (skill: SkillState) => void;
onSkillUpdate: (name: string, updates: Partial<SkillState>) => void;
onFileUpdate: (skillName: string, filename: string, updates: Partial<FileState>) => void;
/** Called when a hunk analysis starts (one SDK invocation per hunk) */
onHunkStart?: (skillName: string, filename: string, hunkNum: number, totalHunks: number, lineRange: string) => void;
onChunkComplete?: (skillName: string, chunk: ChunkAnalysisResult) => void;
onSkillComplete: (name: string, report: SkillReport) => void;
onSkillSkipped: (name: string) => void;
onSkillError: (name: string, error: string) => void;
/** Called when a prompt exceeds the large prompt threshold */
onLargePrompt?: (skillName: string, filename: string, lineRange: string, chars: number, estimatedTokens: number) => void;
/** Called with prompt size info in debug mode */
onPromptSize?: (skillName: string, filename: string, lineRange: string, systemChars: number, userChars: number, totalChars: number, estimatedTokens: number) => void;
/** Called with extraction result details (debug mode) */
onExtractionResult?: (skillName: string, filename: string, lineRange: string, findingsCount: number, method: 'regex' | 'llm' | 'none') => void;
/** Called when findings are dropped, revised, merged, or stripped after analysis */
onFindingProcessing?: (skillName: string, event: FindingProcessingEvent) => void;
/** Called when hunk analysis fails (SDK error, API error, abort) */
onHunkFailed?: (skillName: string, filename: string, lineRange: string, error: string) => void;
/** Called when findings extraction fails (both regex and LLM fallback failed) */
onExtractionFailure?: (skillName: string, filename: string, lineRange: string, error: string, preview: string) => void;
/** Called when a retry attempt is made */
onRetry?: (skillName: string, filename: string, lineRange: string, attempt: number, maxRetries: number, error: string, delayMs: number) => void;
}
/**
* Run a single skill task.
*/
export async function runSkillTask(
options: SkillTaskOptions,
fileConcurrency: number,
callbacks: SkillProgressCallbacks,
semaphore?: Semaphore
): Promise<SkillTaskResult> {
const { name, displayName = name, triggerName, failOn, minConfidence, resolveSkill, context, runnerOptions = {} } = options;
return Sentry.startSpan(
{ op: 'skill.run', name: `run ${displayName}` },
async (span) => {
span.setAttribute('gen_ai.agent.name', displayName);
Iif (triggerName) {
span.setAttribute('warden.trigger.name', triggerName);
}
const files = context.pullRequest?.files ?? [];
span.setAttribute('warden.file.count', files.length);
logger.info(logger.fmt`Skill execution started: ${displayName}`, {
'warden.file.count': files.length,
});
const startTime = Date.now();
const runtime = runnerOptions.runtime ?? 'pi';
// Mirror of the inner-scope `skill` so the outer catch can use
// report.skill when resolveSkill succeeded but a later step threw.
// Stays undefined only if resolveSkill itself failed.
let resolvedSkillName: string | undefined;
try {
let skill: SkillDefinition;
try {
skill = await resolveSkill();
resolvedSkillName = skill.name;
span.setAttribute('gen_ai.agent.name', skill.name);
} catch (err) {
if (err instanceof WardenAuthenticationError) throw err;
const message = err instanceof Error ? err.message : String(err);
throw new SkillRunnerError(message, { cause: err, code: 'skill_resolution_failed' });
}
// Prepare files (parse patches into hunks)
const { files: preparedFiles, skippedFiles } = prepareFiles(context, {
contextLines: runnerOptions.contextLines,
});
if (preparedFiles.length === 0) {
// No files to analyze - skip
const skippedReport: SkillReport = {
skill: skill.name,
summary: 'No code changes to analyze',
findings: [],
usage: { inputTokens: 0, outputTokens: 0, costUSD: 0 },
durationMs: Date.now() - startTime,
model: runnerOptions?.model,
runtime,
skippedFiles: skippedFiles.length > 0 ? skippedFiles : undefined,
};
span.setAttribute('warden.finding.count', 0);
callbacks.onSkillSkipped(name);
// Also fire onSkillComplete so the incremental JSONL writer records the skipped skill.
callbacks.onSkillComplete(name, skippedReport);
return {
name,
report: skippedReport,
failOn,
minConfidence,
};
}
// Initialize file states
const fileStates: FileState[] = preparedFiles.map((file) => ({
filename: file.filename,
status: 'pending',
currentHunk: 0,
totalHunks: file.hunks.length,
findings: [],
}));
// Notify skill start
callbacks.onSkillStart({
name,
displayName,
status: 'running',
startTime,
files: fileStates,
findings: [],
});
// Build PR context for inclusion in prompts (if available)
// For non-PR contexts (CLI file/diff mode), skip the "Other Files" list to avoid
// bloating every hunk prompt with thousands of filenames.
const isPullRequest = context.pullRequest ? context.pullRequest.number !== 0 : false;
const prContext: PRPromptContext | undefined = context.pullRequest
? {
changedFiles: isPullRequest ? context.pullRequest.files.map((f) => f.filename) : [],
title: context.pullRequest.title,
body: context.pullRequest.body,
maxContextFiles: runnerOptions.maxContextFiles,
}
: undefined;
// Process files with concurrency
const processFile = async (prepared: PreparedFile, index: number): Promise<FileProcessResult> => {
const filename = prepared.filename;
const fileStartTime = Date.now();
// Update file state to running (local + callback)
const localState = fileStates[index];
Eif (localState) localState.status = 'running';
callbacks.onFileUpdate(name, filename, { status: 'running' });
const fileCallbacks: FileAnalysisCallbacks = {
skillStartTime: startTime,
onHunkStart: (hunkNum, totalHunks, lineRange) => {
callbacks.onFileUpdate(name, filename, {
currentHunk: hunkNum,
totalHunks,
});
callbacks.onHunkStart?.(name, filename, hunkNum, totalHunks, lineRange);
},
onHunkComplete: (_hunkNum, findings, usage) => {
// Accumulate findings and usage for this file
const current = fileStates[index];
if (current) {
current.findings.push(...findings);
if (current.usage) {
current.usage.inputTokens += usage.inputTokens;
current.usage.outputTokens += usage.outputTokens;
current.usage.costUSD += usage.costUSD;
if (usage.cacheReadInputTokens) {
current.usage.cacheReadInputTokens = (current.usage.cacheReadInputTokens ?? 0) + usage.cacheReadInputTokens;
}
if (usage.cacheCreationInputTokens) {
current.usage.cacheCreationInputTokens = (current.usage.cacheCreationInputTokens ?? 0) + usage.cacheCreationInputTokens;
}
if (usage.cacheCreation5mInputTokens) {
current.usage.cacheCreation5mInputTokens = (current.usage.cacheCreation5mInputTokens ?? 0) + usage.cacheCreation5mInputTokens;
}
if (usage.cacheCreation1hInputTokens) {
current.usage.cacheCreation1hInputTokens = (current.usage.cacheCreation1hInputTokens ?? 0) + usage.cacheCreation1hInputTokens;
}
if (usage.webSearchRequests) {
current.usage.webSearchRequests = (current.usage.webSearchRequests ?? 0) + usage.webSearchRequests;
}
} else {
current.usage = { ...usage };
}
callbacks.onFileUpdate(name, filename, { usage: current.usage });
}
},
onLargePrompt: callbacks.onLargePrompt
? (lineRange, chars, estimatedTokens) => {
callbacks.onLargePrompt?.(name, filename, lineRange, chars, estimatedTokens);
}
: undefined,
onPromptSize: callbacks.onPromptSize
? (lineRange, systemChars, userChars, totalChars, estimatedTokens) => {
callbacks.onPromptSize?.(name, filename, lineRange, systemChars, userChars, totalChars, estimatedTokens);
}
: undefined,
onExtractionResult: callbacks.onExtractionResult
? (lineRange, findingsCount, method) => {
callbacks.onExtractionResult?.(name, filename, lineRange, findingsCount, method);
}
: undefined,
onChunkComplete: callbacks.onChunkComplete
? (chunk) => {
callbacks.onChunkComplete?.(skill.name, chunk);
}
: undefined,
onHunkFailed: callbacks.onHunkFailed
? (lineRange, error) => {
callbacks.onHunkFailed?.(name, filename, lineRange, error);
}
: undefined,
onExtractionFailure: callbacks.onExtractionFailure
? (lineRange, error, preview) => {
callbacks.onExtractionFailure?.(name, filename, lineRange, error, preview);
}
: undefined,
onRetry: callbacks.onRetry
? (lineRange, attempt, maxRetries, error, delayMs) => {
callbacks.onRetry?.(name, filename, lineRange, attempt, maxRetries, error, delayMs);
}
: undefined,
};
const result = await analyzeFile(
skill,
prepared,
context.repoPath,
runnerOptions,
fileCallbacks,
prContext
);
// Detect if this file was aborted before any real work happened
const fileDurationMs = Date.now() - fileStartTime;
const aborted = runnerOptions.abortController?.signal.aborted ?? false;
const noWork = !result.usage || (result.usage.inputTokens === 0 && result.usage.outputTokens === 0);
const fileStatus = (aborted && noWork) ? 'skipped' : 'done';
Eif (localState) localState.status = fileStatus;
callbacks.onFileUpdate(name, filename, {
status: fileStatus,
findings: result.findings,
usage: result.usage,
durationMs: fileDurationMs,
});
return {
findings: result.findings,
usage: result.usage,
durationMs: fileDurationMs,
failedHunks: result.failedHunks,
failedExtractions: result.failedExtractions,
hunkFailures: result.hunkFailures,
auxiliaryUsage: result.auxiliaryUsage,
};
};
// Return an empty result for files skipped due to abort
const processSkippedFile = (index: number): FileProcessResult => {
const localState = fileStates[index];
if (localState) localState.status = 'skipped';
const filename = preparedFiles[index]?.filename ?? 'unknown';
callbacks.onFileUpdate(name, filename, { status: 'skipped' });
return { findings: [], durationMs: 0, failedHunks: 0, failedExtractions: 0, hunkFailures: [] };
};
// Process files with sliding-window concurrency pool
const batchDelayMs = runnerOptions.batchDelayMs ?? 0;
const shouldAbort = () => runnerOptions.abortController?.signal.aborted ?? false;
// The effective concurrency for batch delay: when a semaphore gates work,
// use its permit count (the actual concurrency limit) rather than fileConcurrency.
const effectiveConcurrency = semaphore ? semaphore.initialPermits : fileConcurrency;
const allResults = await runPool(preparedFiles, fileConcurrency,
async (file, index) => {
if (semaphore) await semaphore.acquire();
try {
// Check abort after acquiring the semaphore -- the file may have
// been queued behind others and a SIGINT could have arrived while waiting.
Iif (shouldAbort()) return processSkippedFile(index);
// Rate-limit: delay items beyond the first concurrent wave
Iif (index >= effectiveConcurrency && batchDelayMs > 0) {
await new Promise((resolve) => setTimeout(resolve, batchDelayMs));
}
return await processFile(file, index);
} finally {
if (semaphore) semaphore.release();
}
},
{ shouldAbort }
);
// Mark never-dispatched files as skipped
for (const fileState of fileStates) {
Iif (fileState.status === 'pending') {
callbacks.onFileUpdate(name, fileState.filename, { status: 'skipped' });
}
}
// Build report
const duration = Date.now() - startTime;
const allFindings = allResults.flatMap((r) => r.findings);
const allUsage = allResults.map((r) => r.usage).filter((u): u is UsageStats => u !== undefined);
const allAuxEntries = allResults.flatMap((r) => r.auxiliaryUsage ?? []);
const totalFailedHunks = allResults.reduce((sum, r) => sum + r.failedHunks, 0);
const totalFailedExtractions = allResults.reduce((sum, r) => sum + r.failedExtractions, 0);
const allHunkFailures: HunkFailure[] = allResults.flatMap((r) => r.hunkFailures);
const totalHunks = preparedFiles.reduce((sum, f) => sum + f.hunks.length, 0);
// Each hunk contributes to at most one of failedHunks / failedExtractions
// (mutually exclusive in analyzeFile), so summing them gives the total
// failed-hunk count. Counting only analysis failures would miss the
// scenario where every hunk's SDK call succeeded but every extraction
// failed — a silent zero-findings run otherwise.
const totalAttemptFailures = totalFailedHunks + totalFailedExtractions;
const circuitReason = runnerOptions.circuitBreaker?.reason;
if (
totalHunks > 0
&& allFindings.length === 0
&& totalAttemptFailures > 0
&& (
circuitReason
|| (
totalAttemptFailures === totalHunks
&& !(runnerOptions.abortController?.signal.aborted ?? false)
)
)
) {
const auxUsage = aggregateAuxiliaryUsage(allAuxEntries);
const error = summarizeRunFailure({
totalHunks,
hunkFailures: allHunkFailures,
circuitReason,
runtime: runnerOptions.runtime,
});
const errorReport: SkillReport = {
skill: skill.name,
summary: `${skill.name}: failed (${error.code})`,
findings: [],
usage: aggregateUsage(allUsage),
durationMs: duration,
model: runnerOptions?.model,
runtime,
// Preserve per-file metadata (timing, partial usage, attempted
// filenames) on failure runs too — `warden runs` and JSONL
// consumers iterate this array to count attempted files. Without
// it, a failed run shows totalFiles: 0.
files: preparedFiles.map((file, i) => {
const r = allResults[i];
return {
filename: file.filename,
findings: r?.findings.length ?? 0,
durationMs: r?.durationMs,
usage: r?.usage,
};
}),
failedHunks: totalFailedHunks,
hunkFailures: allHunkFailures,
error: {
code: error.code,
message: error.message,
timestamp: new Date().toISOString(),
},
};
if (totalFailedExtractions > 0) errorReport.failedExtractions = totalFailedExtractions;
Iif (skippedFiles.length > 0) errorReport.skippedFiles = skippedFiles;
Iif (auxUsage) errorReport.auxiliaryUsage = auxUsage;
span.setAttribute('warden.finding.count', 0);
callbacks.onSkillError(name, error.message);
// Mirror the success path: emit a final completion event with the
// (errored) report so terminal renderers print the per-skill
// summary line. Without this, console mode shows the error string
// alone with no breakdown of timing, cost, or attempted files.
callbacks.onSkillUpdate(name, {
status: 'error',
durationMs: duration,
findings: [],
usage: errorReport.usage,
auxiliaryUsage: errorReport.auxiliaryUsage,
});
callbacks.onSkillComplete(name, errorReport);
// Carry a typed error alongside the report so consumers that re-throw
// (action executor, Sentry.captureException) preserve the ErrorCode.
const runnerError = new SkillRunnerError(error.message, { code: error.code });
return { name, report: errorReport, error: runnerError, failOn, minConfidence };
}
const processed = await postProcessFindings(allFindings, {
skill,
repoPath: context.repoPath,
apiKey: runnerOptions.apiKey,
runtime: runnerOptions.runtime,
auxiliaryModel: runnerOptions.auxiliaryModel,
synthesisModel: runnerOptions.synthesisModel,
auxiliaryMaxRetries: runnerOptions.auxiliaryMaxRetries,
verifyFindings: runnerOptions.verifyFindings,
maxTurns: runnerOptions.maxTurns,
abortController: runnerOptions.abortController,
pathToClaudeCodeExecutable: runnerOptions.pathToClaudeCodeExecutable,
prContext,
onFindingProcessing: (event) => {
callbacks.onFindingProcessing?.(name, event);
},
});
const finalFindings = processed.findings;
allAuxEntries.push(...processed.auxiliaryUsage);
const report: SkillReport = {
skill: skill.name,
summary: generateSummary(skill.name, finalFindings),
findings: finalFindings,
usage: aggregateUsage(allUsage),
durationMs: duration,
model: runnerOptions?.model,
runtime,
files: buildFileReports(
preparedFiles.map((file, i) => {
const r = allResults[i];
return {
filename: file.filename,
durationMs: r?.durationMs,
usage: r?.usage,
};
}),
finalFindings,
),
};
Iif (skippedFiles.length > 0) {
report.skippedFiles = skippedFiles;
}
if (totalFailedHunks > 0) {
report.failedHunks = totalFailedHunks;
}
Iif (totalFailedExtractions > 0) {
report.failedExtractions = totalFailedExtractions;
}
if (allHunkFailures.length > 0) {
report.hunkFailures = allHunkFailures;
}
const auxUsage = aggregateAuxiliaryUsage(allAuxEntries);
Iif (auxUsage) {
report.auxiliaryUsage = auxUsage;
}
span.setAttribute('warden.finding.count', report.findings.length);
// Emit metrics and log completion
emitSkillMetrics(report);
logger.info(logger.fmt`Skill execution complete: ${displayName}`, {
'warden.finding.count': report.findings.length,
'duration_ms': report.durationMs,
});
// Notify skill complete
callbacks.onSkillUpdate(name, {
status: 'done',
durationMs: duration,
findings: finalFindings,
usage: report.usage,
auxiliaryUsage: report.auxiliaryUsage,
});
callbacks.onSkillComplete(name, report);
return { name, report, failOn, minConfidence };
} catch (err) {
const { code, message } = classifyError(err);
callbacks.onSkillError(name, message);
// Use the resolved skill name when available so JSONL output matches
// the success path's identifier. Falls back to the trigger name only
// when resolveSkill itself threw.
const skillName = resolvedSkillName ?? name;
const errorReport: SkillReport = {
skill: skillName,
summary: `${skillName}: failed (${code})`,
findings: [],
durationMs: Date.now() - startTime,
model: runnerOptions?.model,
runtime,
error: { code, message, timestamp: new Date().toISOString() },
};
span.setAttribute('warden.finding.count', 0);
// Mirror the success / all-hunks-fail paths: emit a final completion
// event so non-TTY (log-mode) renderers print a per-skill summary
// line for the failure. Without this, log mode shows only the
// bare error string with no timing or duration.
callbacks.onSkillUpdate(name, {
status: 'error',
durationMs: errorReport.durationMs,
findings: [],
});
callbacks.onSkillComplete(name, errorReport);
return { name, report: errorReport, error: err, failOn, minConfidence };
}
},
);
}
/**
* Create default progress callbacks for console output.
* In TTY mode: colored icons, chalk formatting.
* In non-TTY/log mode: timestamped lines with finding details.
*/
export function createDefaultCallbacks(
tasks: SkillTaskOptions[],
mode: OutputMode,
verbosity: Verbosity
): SkillProgressCallbacks {
/** Resolve the display name for a skill, falling back to the raw name. */
function displayNameFor(name: string): string {
return tasks.find((t) => t.name === name)?.displayName ?? name;
}
/** Track per-skill skipped file counts for collapsed summary in non-TTY mode. */
const skippedCounts = new Map<string, number>();
// Skipped skills also fire onSkillComplete (for the JSONL writer).
// Suppress the duplicate "completed" line for those names.
const skippedSkills = new Set<string>();
return {
onSkillStart: (skill) => {
if (verbosity === Verbosity.Quiet) return;
if (!mode.isTTY) {
const fileCount = skill.files.length;
logPlain(`Running ${displayNameFor(skill.name)} (${fileCount} ${pluralize(fileCount, 'file')})...`);
}
},
onSkillUpdate: () => { /* no-op for default callbacks */ },
onFileUpdate: (_skillName, filename, updates) => {
if (updates.status === 'skipped') {
skippedCounts.set(_skillName, (skippedCounts.get(_skillName) ?? 0) + 1);
return;
}
if (verbosity === Verbosity.Quiet || mode.isTTY) return;
if (updates.status !== 'done') return;
const duration = updates.durationMs !== undefined ? formatDuration(updates.durationMs) : '?';
const cost = updates.usage ? ` ${formatCost(updates.usage.costUSD)}` : '';
const n = updates.findings?.length ?? 0;
const suffix = n > 0 ? ` ${n} ${pluralize(n, 'finding')}` : '';
logPlain(` ${displayNameFor(_skillName)} > ${filename} done ${duration}${cost}${suffix}`);
},
onHunkStart: (skillName, filename, hunkNum, totalHunks, lineRange) => {
if (verbosity === Verbosity.Quiet || mode.isTTY) return;
logPlain(` ${displayNameFor(skillName)} > ${filename} [${hunkNum}/${totalHunks}] ${lineRange}`);
},
onSkillComplete: (name, report) => {
if (verbosity === Verbosity.Quiet) return;
if (skippedSkills.has(name)) return;
const displayName = displayNameFor(name);
// Errored runs render as failures, not as misleading "completed -
// 0 findings" lines with a green checkmark. onSkillError already
// printed the error message; this line carries timing only.
Iif (report.error) {
if (mode.isTTY) {
const duration = report.durationMs !== undefined ? ` ${chalk.dim(`[${formatDuration(report.durationMs)}]`)}` : '';
console.error(`${chalk.red('✗')} ${displayName}${duration} ${chalk.red(`(${report.error.code})`)}`);
} else {
const duration = report.durationMs !== undefined ? formatDuration(report.durationMs) : '?';
logPlain(`${displayName} failed in ${duration} (${report.error.code})`);
}
return;
}
Iif (mode.isTTY) {
const duration = report.durationMs !== undefined ? ` ${chalk.dim(`[${formatDuration(report.durationMs)}]`)}` : '';
console.error(`${chalk.green(ICON_CHECK)} ${displayName}${duration}`);
// Debug: log finding details
if (verbosity >= Verbosity.Debug && report.findings.length > 0) {
for (const finding of report.findings) {
debugLog(mode, `${formatSeverityPlain(finding.severity)} ${findingLocation(finding)}: ${finding.title}`);
if (finding.suggestedFix) {
debugLog(mode, ` fix: ${finding.suggestedFix.description}`);
}
}
}
} else {
// Log mode: timestamped completion with duration and finding summary
const duration = report.durationMs !== undefined ? formatDuration(report.durationMs) : '?';
const counts = countBySeverity(report.findings);
const summary = formatFindingCountsPlain(counts);
logPlain(`${displayName} completed in ${duration} - ${summary}`);
// Show per-finding lines at Verbose+ verbosity in log mode
// (the final report already shows findings with full detail)
if (verbosity >= Verbosity.Verbose) {
for (const finding of report.findings) {
logPlain(` ${formatSeverityPlain(finding.severity)} ${findingLocation(finding)}: ${finding.title}`);
if (verbosity >= Verbosity.Debug && finding.suggestedFix) {
logPlain(` fix: ${finding.suggestedFix.description}`);
}
}
}
const skipped = skippedCounts.get(name) ?? 0;
if (skipped > 0) {
logPlain(` ${skipped} ${pluralize(skipped, 'file')} skipped`);
}
}
},
onSkillSkipped: (name) => {
skippedSkills.add(name);
Iif (verbosity === Verbosity.Quiet) return;
const displayName = displayNameFor(name);
Iif (mode.isTTY) {
console.error(`${chalk.yellow(ICON_SKIPPED)} ${displayName} ${chalk.dim('[skipped]')}`);
} else {
logPlain(`${displayName} skipped`);
}
},
onSkillError: (name, error) => {
Iif (verbosity === Verbosity.Quiet) return;
const displayName = displayNameFor(name);
Iif (mode.isTTY) {
console.error(`${chalk.red('\u2717')} ${displayName} - ${chalk.red(error)}`);
} else {
logPlain(`ERROR: ${displayName} - ${error}`);
const skipped = skippedCounts.get(name) ?? 0;
if (skipped > 0) {
logPlain(` ${skipped} ${pluralize(skipped, 'file')} skipped`);
}
}
},
// Warn about large prompts (always shown unless quiet)
onLargePrompt: (_skillName, filename, lineRange, chars, estimatedTokens) => {
Iif (verbosity === Verbosity.Quiet) return;
const location = `${filename}:${lineRange}`;
const size = `${Math.round(chars / 1000)}k chars (~${Math.round(estimatedTokens / 1000)}k tokens)`;
Iif (mode.isTTY) {
console.error(`${chalk.yellow(figures.warning)} Large prompt for ${location}: ${size}`);
} else {
logPlain(`WARN: Large prompt for ${location}: ${size}`);
}
},
// Debug mode: show prompt sizes
onPromptSize: verbosity >= Verbosity.Debug
? (_skillName, filename, lineRange, systemChars, userChars, totalChars, estimatedTokens) => {
const location = `${filename}:${lineRange}`;
debugLog(mode, `Prompt for ${location}: system=${systemChars}, user=${userChars}, total=${totalChars} chars (~${estimatedTokens} tokens)`);
}
: undefined,
// Debug mode: show extraction results
onExtractionResult: verbosity >= Verbosity.Debug
? (_skillName, filename, lineRange, findingsCount, method) => {
debugLog(mode, `Extracted ${findingsCount} ${pluralize(findingsCount, 'finding')} from ${filename}:${lineRange} via ${method}`);
}
: undefined,
onFindingProcessing: verbosity >= Verbosity.Debug
? (_skillName, event) => {
debugLog(mode, formatFindingProcessingEvent(event));
}
: undefined,
// Verbose mode: show per-hunk analysis failures (spec: event #16 hunk_failed)
onHunkFailed: verbosity >= Verbosity.Verbose
? (_skillName, filename, lineRange, error) => {
const location = `${filename}:${lineRange}`;
if (mode.isTTY) {
console.error(`${chalk.yellow(figures.warning)} Chunk failed: ${location} ${chalk.dim(`\u2014 ${error}`)}`);
} else {
logPlain(`WARN: Chunk failed: ${location} \u2014 ${error}`);
}
}
: undefined,
// Verbose mode: show per-hunk extraction failures (spec: event #17 extraction_failure)
onExtractionFailure: verbosity >= Verbosity.Verbose
? (_skillName, filename, lineRange, error, preview) => {
const location = `${filename}:${lineRange}`;
Iif (mode.isTTY) {
console.error(`${chalk.yellow(figures.warning)} Extraction failed: ${location} ${chalk.dim(`\u2014 ${error}`)}`);
if (verbosity >= Verbosity.Debug && preview) {
debugLog(mode, ` Output preview: ${preview.slice(0, 200)}`);
}
} else {
logPlain(`WARN: Extraction failed: ${location} \u2014 ${error}`);
if (verbosity >= Verbosity.Debug && preview) {
logPlain(`DEBUG: Output preview: ${preview.slice(0, 200)}`);
}
}
}
: undefined,
// Verbose mode: show retry attempts (spec: event #18 retry)
onRetry: verbosity >= Verbosity.Verbose
? (_skillName, filename, lineRange, attempt, maxRetries, error, delayMs) => {
const location = `${filename}:${lineRange}`;
const retryInfo = `attempt ${attempt}/${maxRetries}`;
const delay = delayMs > 0 ? `, retrying in ${Math.round(delayMs / 1000)}s` : '';
Iif (mode.isTTY) {
debugLog(mode, `Retry ${location} (${retryInfo}${delay}): ${error}`);
} else {
logPlain(`WARN: Retry ${location} (${retryInfo}${delay}): ${error}`);
}
}
: undefined,
};
}
function composeAbortControllers(...controllers: (AbortController | undefined)[]): AbortController {
const composed = new AbortController();
for (const ctrl of controllers) {
if (ctrl?.signal.aborted) {
composed.abort();
return composed;
}
ctrl?.signal.addEventListener('abort', () => composed.abort(), { once: true });
}
return composed;
}
/**
* Share abort/circuit state across task runner options.
*/
export function composeTasksWithFailFast(
tasks: SkillTaskOptions[],
failFastController?: AbortController,
circuitBreaker?: ProviderFailureCircuitBreaker,
circuitAbortController?: AbortController,
): SkillTaskOptions[] {
Iif (!failFastController && !circuitBreaker && !circuitAbortController) return tasks;
return tasks.map((task) => ({
...task,
runnerOptions: {
...task.runnerOptions,
abortController: composeAbortControllers(
task.runnerOptions?.abortController,
failFastController,
circuitAbortController,
),
circuitBreaker: task.runnerOptions?.circuitBreaker ?? circuitBreaker,
},
}));
}
/**
* Launch all skill tasks in parallel using a shared semaphore for concurrency.
*/
export async function runComposedSkillTasks(
tasks: SkillTaskOptions[],
callbacks: SkillProgressCallbacks,
semaphore: Semaphore
): Promise<SkillTaskResult[]> {
const results = await runPool(tasks, tasks.length,
(task) => runSkillTask(task, Number.MAX_SAFE_INTEGER, callbacks, semaphore),
{ shouldAbort: () => tasks[0]?.runnerOptions?.abortController?.signal.aborted ?? false }
);
return results;
}
/**
* Run multiple skill tasks with optional concurrency.
* Uses callbacks to report progress for Ink rendering.
*/
export async function runSkillTasks(
tasks: SkillTaskOptions[],
options: RunTasksOptions,
callbacks?: SkillProgressCallbacks
): Promise<SkillTaskResult[]> {
const { mode, verbosity, concurrency, failFastController, onSkillComplete, onChunkComplete } = options;
// Global semaphore gates file-level work across all skills.
// All skills launch immediately so the UI shows them as "running",
// but only `concurrency` files will be analysed at any time.
const semaphore = new Semaphore(concurrency);
const effectiveCallbacks = callbacks ?? createDefaultCallbacks(tasks, mode, verbosity);
const wrappedCallbacks: SkillProgressCallbacks = {
...effectiveCallbacks,
...(onSkillComplete || failFastController
? {
onSkillComplete: (name: string, report: SkillReport) => {
effectiveCallbacks.onSkillComplete(name, report);
try { onSkillComplete?.(report); } catch { /* streaming hook must not break the run */ }
if (failFastController && report.findings.length > 0) {
failFastController.abort();
}
},
}
: {}),
...(onChunkComplete
? {
onChunkComplete: (name: string, chunk: ChunkAnalysisResult) => {
effectiveCallbacks.onChunkComplete?.(name, chunk);
try { onChunkComplete(name, chunk); } catch { /* streaming hook must not break the run */ }
},
}
: {}),
};
// Output SKILLS header (TTY only - in log mode, "Running..." lines are sufficient)
Iif (verbosity !== Verbosity.Quiet && tasks.length > 0 && mode.isTTY) {
console.error(chalk.bold('SKILLS'));
}
const circuitAbortController = new AbortController();
const circuitBreaker = new ProviderFailureCircuitBreaker({ abortController: circuitAbortController });
const composedTasks = composeTasksWithFailFast(
tasks,
failFastController,
circuitBreaker,
circuitAbortController,
);
// Listen for abort signal to show interrupt message (non-TTY only; Ink handles TTY)
const abortSignal = composedTasks[0]?.runnerOptions?.abortController?.signal;
Iif (abortSignal && !abortSignal.aborted && !mode.isTTY && verbosity !== Verbosity.Quiet) {
abortSignal.addEventListener('abort', () => {
// Only show interrupt message for user SIGINT, not fail-fast
if (!failFastController?.signal.aborted && !circuitAbortController.signal.aborted) {
logPlain('Interrupted, finishing up... (press Ctrl+C again to force exit)');
}
}, { once: true });
}
// Show fail-fast message when triggered (non-TTY only)
Iif (failFastController && !mode.isTTY && verbosity !== Verbosity.Quiet) {
failFastController.signal.addEventListener('abort', () => {
logPlain('Stopping \u2014 finding detected (--fail-fast)');
}, { once: true });
}
// Launch all skills in parallel; the semaphore is the sole concurrency gate.
return runComposedSkillTasks(composedTasks, wrappedCallbacks, semaphore);
}
|