TypeScript AI agent orchestration framework with dynamic workflows. Describe the goal, not the graph: a coordinator plans the task DAG at runtime and runs it on any LLM (Claude, ChatGPT, Gemini, DeepSeek, or local models).
This report presents the forensic synthetic code analysis of open-multi-agent/open-multi-agent, a TypeScript project with 6,580 GitHub stars. SynthScan v2.0 examined 74,994 lines of code across 260 source files, recording 200 pattern matches distributed across 8 syntactic categories. The overall adjusted score of 4.0 places this repository in the Likely human-written band.
The scanner applied 160+ deterministic lexical heuristics, multi-line block detectors, abstract syntax tree depth profilers, and a cross-file Jaccard similarity matrix to construct a statistically normalised synthetic code estimate. All matches are individually weighted by severity coefficient and contextual multiplier before summation, and the resulting headline score is temporally discounted to account for the repository's development history relative to the commercial emergence of large language model coding tooling (November 2022 onward).
This chart maps the temporal evolution of the adjusted synthetic code score across successive scan runs. An upward trajectory indicates ongoing incorporation of AI-generated code or expanding LLM-assisted scaffolding; a stable or declining trajectory may reflect active human refactoring, code removal, or the adoption of stricter authorship policies. The dashed secondary line (right axis) independently tracks total raw pattern hit count, which can diverge from the normalised score when codebase size changes significantly between scans.
Classifies detected patterns by their diagnostic confidence and structural impact. CRITICAL patterns (coefficient 10) represent definitive synthetic signatures — hallucinated imports, explicit LLM attribution metadata — virtually never produced by human authors. HIGH (5) indicates strong structural tells such as cross-file repetition or cross-linguistic idioms. MEDIUM (2) covers recognisable conversational padding and AI-specific vocabulary. LOW (1) captures subtle indicators like tautological comments and generic boilerplate that require density to carry independent signal.
This horizontal bar chart decomposes the repository's raw synthetic code score by top-level directory, allowing you to pinpoint precisely which modules or components carry the highest AI authorship density. Directories with disproportionately high scores relative to their size warrant targeted manual review: concentrated AI signatures often trace back to mass-generated configuration layers, auto-ported test suites, LLM-scaffolded boilerplate classes, or entire subsystems authored under heavy copilot assistance. Use this view to prioritise your human code-review effort.
The scanner identified 200 distinct pattern matches across 8 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a statistically significant AI authorship indicator. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.
Reading the findings table: The Severity column indicates the diagnostic confidence level (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — comment-context matches receive a ×1.5 weight because LLMs systematically over-annotate. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier as dense clusters constitute far stronger evidence of synthetic authorship than isolated hits.
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | packages/core/examples/basics/multi-model-team.ts | 185 | // Step 1: researcher fetches the rates | COMMENT |
| LOW | packages/core/examples/basics/multi-model-team.ts | 205 | // Step 2: analyst writes the briefing, receiving the researcher output as context | COMMENT |
| LOW | packages/core/examples/patterns/fan-out-aggregate.ts | 118 | // Step 1: Fan-out — run all 3 analysts in parallel | COMMENT |
| LOW | packages/core/examples/patterns/fan-out-aggregate.ts | 148 | // Step 2: Aggregate — synthesizer reads all 3 analyses | COMMENT |
| LOW | …ges/core/examples/patterns/cross-provider-reasoning.ts | 53 | // Step 1: a reasoning-capable OpenAI model produces a turn that includes | COMMENT |
| LOW | …ges/core/examples/patterns/cross-provider-reasoning.ts | 76 | // Step 2: swap the model to Anthropic mid-conversation. Without | COMMENT |
| LOW⚡ | packages/core/examples/cookbook/contract-review-dag.ts | 46 | // Step 1: Agent configurations (4 agents) | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 127 | // Step 2: Task configurations (4 tasks) | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 143 | // Step 3: Progress tracking | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 218 | // Step 4: Parallelism verification | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 247 | // Step 5: Orchestrator and Team configuration | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 263 | // Step 6: Read contract text | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 317 | // Step 7: Run the DAG | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 155 | // Step 1: Parallel fan-out with per-agent timing | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 205 | // Step 2: Parallelism assertion | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 223 | // Step 3: Aggregate into Markdown report | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 229 | // Step 1: Agent A translates | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 248 | // Step 2: Agent B back-translates | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 265 | // Step 3: Agent C reviews semantic drift | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 42 | // Step 1: Agent configurations (5 agents) | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 153 | // Step 2: Task configurations (5 tasks) | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 169 | // Step 3: Progress tracking | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 244 | // Step 4: Parallelism verification | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 275 | // Step 5: Orchestrator and Team configuration | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 314 | // Step 6: Read fixture files | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 373 | // Step 7: Run the DAG | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 122 | // Step 1: recall long-term memory from previous runs | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 199 | // Step 2: run the team — shared-memory writes are captured into TDAM | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 239 | // Step 3: flush extraction so captures become searchable | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 253 | // Step 4: verify — search the distilled long-term memories | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 167 | // Step 1: Research | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 175 | // Step 2: Fact-check | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 184 | // Step 3: Write briefing | COMMENT |
| LOW | packages/core/src/llm/copilot.ts | 111 | // Step 1: Request a device code | COMMENT |
| LOW⚡ | packages/core/src/llm/copilot.ts | 128 | // Step 2: Prompt the user via callback | COMMENT |
| LOW⚡ | packages/core/src/llm/copilot.ts | 131 | // Step 3: Poll for the user to complete auth | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 883 | // Step 1: Call the LLM and collect the full response for this turn. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 909 | // Step 2: Build the assistant message from the response content. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 995 | // Step 3: Decide whether to continue looping. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 1032 | // Step 4: Execute all tool calls in PARALLEL. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 1137 | // Step 5: Accumulate results and build the user message that carries | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2093 | // Step 1: Coordinator decomposes goal into tasks | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2153 | // Step 2: Parse tasks from coordinator output | COMMENT |
| LOW⚡ | packages/core/src/orchestrator/orchestrator.ts | 2179 | // Step 3: Auto-assign any unassigned tasks | COMMENT |
| LOW⚡ | packages/core/src/orchestrator/orchestrator.ts | 2184 | // Step 4: Build pool and execute | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2297 | // Step 5: Coordinator synthesises final result | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2612 | // Step 2: run proposer(s); accumulate usage and honour the budget before judging. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | packages/core/examples/basics/multi-model-team.ts | 185 | // Step 1: researcher fetches the rates | COMMENT |
| LOW | packages/core/examples/basics/multi-model-team.ts | 205 | // Step 2: analyst writes the briefing, receiving the researcher output as context | COMMENT |
| LOW | packages/core/examples/patterns/fan-out-aggregate.ts | 118 | // Step 1: Fan-out — run all 3 analysts in parallel | COMMENT |
| LOW | packages/core/examples/patterns/fan-out-aggregate.ts | 148 | // Step 2: Aggregate — synthesizer reads all 3 analyses | COMMENT |
| LOW | …ges/core/examples/patterns/cross-provider-reasoning.ts | 53 | // Step 1: a reasoning-capable OpenAI model produces a turn that includes | COMMENT |
| LOW | …ges/core/examples/patterns/cross-provider-reasoning.ts | 76 | // Step 2: swap the model to Anthropic mid-conversation. Without | COMMENT |
| LOW⚡ | packages/core/examples/cookbook/contract-review-dag.ts | 46 | // Step 1: Agent configurations (4 agents) | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 127 | // Step 2: Task configurations (4 tasks) | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 143 | // Step 3: Progress tracking | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 218 | // Step 4: Parallelism verification | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 247 | // Step 5: Orchestrator and Team configuration | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 263 | // Step 6: Read contract text | COMMENT |
| LOW | packages/core/examples/cookbook/contract-review-dag.ts | 317 | // Step 7: Run the DAG | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 155 | // Step 1: Parallel fan-out with per-agent timing | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 205 | // Step 2: Parallelism assertion | COMMENT |
| LOW | packages/core/examples/cookbook/meeting-summarizer.ts | 223 | // Step 3: Aggregate into Markdown report | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 229 | // Step 1: Agent A translates | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 248 | // Step 2: Agent B back-translates | COMMENT |
| LOW | …/core/examples/cookbook/translation-backtranslation.ts | 265 | // Step 3: Agent C reviews semantic drift | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 42 | // Step 1: Agent configurations (5 agents) | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 153 | // Step 2: Task configurations (5 tasks) | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 169 | // Step 3: Progress tracking | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 244 | // Step 4: Parallelism verification | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 275 | // Step 5: Orchestrator and Team configuration | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 314 | // Step 6: Read fixture files | COMMENT |
| LOW | …ages/core/examples/cookbook/incident-postmortem-dag.ts | 373 | // Step 7: Run the DAG | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 122 | // Step 1: recall long-term memory from previous runs | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 199 | // Step 2: run the team — shared-memory writes are captured into TDAM | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 239 | // Step 3: flush extraction so captures become searchable | COMMENT |
| LOW | …integrations/with-tencentdb-memory/team-with-memory.ts | 253 | // Step 4: verify — search the distilled long-term memories | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 167 | // Step 1: Research | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 175 | // Step 2: Fact-check | COMMENT |
| LOW⚡ | …ore/examples/integrations/with-engram/research-team.ts | 184 | // Step 3: Write briefing | COMMENT |
| LOW | packages/core/src/llm/copilot.ts | 111 | // Step 1: Request a device code | COMMENT |
| LOW⚡ | packages/core/src/llm/copilot.ts | 128 | // Step 2: Prompt the user via callback | COMMENT |
| LOW⚡ | packages/core/src/llm/copilot.ts | 131 | // Step 3: Poll for the user to complete auth | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 883 | // Step 1: Call the LLM and collect the full response for this turn. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 909 | // Step 2: Build the assistant message from the response content. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 995 | // Step 3: Decide whether to continue looping. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 1032 | // Step 4: Execute all tool calls in PARALLEL. | COMMENT |
| LOW | packages/core/src/agent/runner.ts | 1137 | // Step 5: Accumulate results and build the user message that carries | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2093 | // Step 1: Coordinator decomposes goal into tasks | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2153 | // Step 2: Parse tasks from coordinator output | COMMENT |
| LOW⚡ | packages/core/src/orchestrator/orchestrator.ts | 2179 | // Step 3: Auto-assign any unassigned tasks | COMMENT |
| LOW⚡ | packages/core/src/orchestrator/orchestrator.ts | 2184 | // Step 4: Build pool and execute | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2297 | // Step 5: Coordinator synthesises final result | COMMENT |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 2612 | // Step 2: run proposer(s); accumulate usage and honour the budget before judging. | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | packages/core/tests/tool-result-compression.test.ts | 54 | function buildErrorRegistryAndExecutor(): { registry: ToolRegistry; executor: ToolExecutor } { | CODE |
| LOW | packages/core/tests/tool-result-compression.test.ts | 70 | function extractToolResultContents(messages: LLMMessage[]): string[] { | CODE |
| LOW | packages/core/tests/checkpoint.test.ts | 116 | async function deleteNonCheckpointEntries(store: MemoryStore): Promise<void> { | CODE |
| LOW | packages/core/tests/structured-output.test.ts | 194 | function buildMockAgentWithAdapter(config: AgentConfig, adapter: LLMAdapter): Agent { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 396 | function extractGithubRepoFullNames(value: unknown): string[] { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 425 | function textContainsNormalizedPhrase(text: string, phrase: string): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 453 | function nonOfficialClaimNearTarget(text: string, target: string | undefined): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 572 | function normalizeDatasetCandidate(raw: string, evidenceKind: DatasetEvidenceKind): string | undefined { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 625 | function datasetEvidenceBaseConfidence(evidenceKind: DatasetEvidenceKind): number { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 640 | function datasetEvidenceConfidence( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 658 | function makeDatasetCandidateEvidence( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 675 | function extractDatasetCandidateEvidenceFromText( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 701 | function extractDatasetCandidateEvidenceFromPaths(paths: string[]): DatasetCandidateEvidence[] { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 725 | function summarizeDatasetCandidates( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 765 | function extractDatasetMentionsFromAstaSnippets(value: unknown): string[] { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 922 | function buildDatasetCluesFromRepos(repoEvidence: Array<Record<string, unknown> | undefined>): Array<Record<string, unkn | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1003 | function isUnavailableSourcePayload(value: unknown): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1033 | function hasSuccessfulSourcePayload(value: unknown): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1047 | function collectSourceEvidenceStrings(value: unknown, acc: string[] = []): string[] { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1073 | function hasReproductionFeedbackSignal(record: Record<string, unknown>): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1078 | function sourceHasReproductionFeedbackSignal(value: unknown): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1095 | function findDatasetRestrictionEvidence(text: string): string | undefined { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1141 | function extractedUnscopedEvidence( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1244 | function liveUnscopedEvidenceFields( | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1317 | function summarizeRelationshipHints(value: unknown): string[] { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1333 | function scoreRepositoryForReplication(repo: Record<string, unknown>): number { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1382 | function compactRepositoryEvidenceForCodeAgent(repo: Record<string, unknown>): Record<string, unknown> { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1408 | function repoHasTrainingOrEvalPath(repo: Record<string, unknown>): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 1427 | function repoHasUnofficialImplementationSignal(repo: Record<string, unknown>): boolean { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 2567 | function assertPlannerDetectedConflicts(plan: ReplicationPlan, mode: SourceMode): void { | CODE |
| LOW | …ges/core/examples/cookbook/paper-replication-triage.ts | 2585 | function verifySourceTasksStartedTogether(resultTasks: readonly TaskExecutionRecord[]): void { | CODE |
| LOW⚡ | packages/core/examples/cookbook/contract-review-dag.ts | 39 | function getAndIncrementAttempt(agentName: string): number { | CODE |
| LOW | packages/core/src/llm/azure-openai.ts | 73 | function resolveAzureDeploymentName(model: string): string { | CODE |
| LOW | packages/core/src/llm/anthropic.ts | 77 | function toAnthropicContentBlockParam( | CODE |
| LOW | packages/core/src/llm/anthropic.ts | 206 | function fromAnthropicContentBlock( | CODE |
| LOW | packages/core/src/llm/reasoning-fallback.ts | 124 | export function resolveReasoningOutboundMaxChars( | CODE |
| LOW | packages/core/src/llm/reasoning-fallback.ts | 176 | export function reasoningBlockToInlineText( | CODE |
| LOW | packages/core/src/llm/ai-sdk.ts | 51 | export function llmMessagesToAiSdkModelMessages( | CODE |
| LOW | packages/core/src/llm/ai-sdk.ts | 174 | function buildLlmResponseFromGenerateText(result: { | CODE |
| LOW | packages/core/src/agent/structured-output.ts | 21 | export function buildStructuredOutputInstruction(schema: ZodSchema): string { | CODE |
| LOW | packages/core/src/agent/runner.ts | 321 | function stripImageBlocksForSummary(messages: LLMMessage[]): LLMMessage[] { | CODE |
| LOW | packages/core/src/agent/runner.ts | 353 | function prependSyntheticPrefixToFirstUser( | CODE |
| LOW | packages/core/src/cli/oma.ts | 353 | async function writeRunTeamDashboardFile(html: string): Promise<string> { | CODE |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 787 | function buildRevealCoordinatorLines( | CODE |
| LOW | packages/core/src/orchestrator/orchestrator.ts | 801 | function prependRevealCoordinatorContext( | CODE |
| LOW | packages/core/src/tool/mcp.ts | 108 | function assertUniqueNormalizedToolNames( | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | packages/core/tests/context-strategy.test.ts | 296 | const FILLER = 'lorem ipsum dolor sit amet '.repeat(50) // ~1.3k chars | CODE |
| LOW | packages/core/tests/context-strategy.test.ts | 296 | const FILLER = 'lorem ipsum dolor sit amet '.repeat(50) // ~1.3k chars | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 58 | ['anthropic', () => new AnthropicAdapter('dummy-key'), 'own-issued' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 59 | ['gemini', () => new GeminiAdapter('dummy-key'), 'own-issued' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 60 | ['openai', () => new OpenAIAdapter('dummy-key'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 61 | ['azure-openai', () => new AzureOpenAIAdapter('dummy-key', 'https://example.openai.azure.com'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 62 | ['copilot', () => new CopilotAdapter('dummy-token'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 63 | ['deepseek', () => new DeepSeekAdapter('dummy-key'), 'tool-use-only' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 64 | ['grok', () => new GrokAdapter('dummy-key'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 65 | ['qiniu', () => new QiniuAdapter('dummy-key'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 66 | ['minimax', () => new MiniMaxAdapter('dummy-key'), 'never' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 67 | ['mimo', () => new MiMoAdapter('dummy-key'), 'tool-use-only' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 68 | ['hunyuan', () => new HunyuanAdapter('dummy-key'), 'tool-use-only' as const], | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 81 | const parent = new OpenAIAdapter('dummy-key').capabilities | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 82 | expect(new GrokAdapter('dummy-key').capabilities).toEqual(parent) | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 83 | expect(new QiniuAdapter('dummy-key').capabilities).toEqual(parent) | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 84 | expect(new MiniMaxAdapter('dummy-key').capabilities).toEqual(parent) | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 93 | expect(new DeepSeekAdapter('dummy-key').capabilities).toEqual({ | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 96 | expect(new MiMoAdapter('dummy-key').capabilities).toEqual({ | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 99 | expect(new HunyuanAdapter('dummy-key').capabilities).toEqual({ | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 102 | expect(new DeepSeekAdapter('dummy-key').capabilities).not.toEqual( | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 103 | new OpenAIAdapter('dummy-key').capabilities, | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 105 | expect(new MiMoAdapter('dummy-key').capabilities).not.toEqual( | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 106 | new OpenAIAdapter('dummy-key').capabilities, | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 108 | expect(new HunyuanAdapter('dummy-key').capabilities).not.toEqual( | CODE |
| LOW⚡ | packages/core/tests/llm-adapter-capabilities.test.ts | 109 | new OpenAIAdapter('dummy-key').capabilities, | CODE |
| LOW | packages/core/tests/hunyuan-adapter.test.ts | 48 | const adapter = new HunyuanAdapter('dummy-key') | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | README.md | 92 | - **[CodingScaffold](https://github.com/JRS1986/CodingScaffold)** — Agentic-coding scaffold that lists OMA as an optiona | CODE |
| MEDIUM | README.md | 111 | │ └── core/ # @open-multi-agent/core — the published library | CODE |
| MEDIUM | README_zh.md | 100 | - **[CodingScaffold](https://github.com/JRS1986/CodingScaffold)** — agentic-coding 脚手架，把 OMA 列为可选编排后端，附带 `runTeam` 工作流模板 | CODE |
| MEDIUM | README_zh.md | 119 | │ └── core/ # @open-multi-agent/core（发布的库） | CODE |
| MEDIUM | docs/checkpoint.md | 180 | These are tracked as follow-ups: [#312](https://github.com/open-multi-agent/open-multi-agent/issues/312) (mid-task recov | CODE |
| MEDIUM | docs/tool-configuration.md | 28 | **Custom / runtime tools are exempt from the grant requirement** — registering them _is_ the grant. Tools passed via `cu | CODE |
| MEDIUM | packages/core/README.md | 59 | [Quick Start](#quick-start) · [Three Ways to Run](#three-ways-to-run) · [Features](#features) · [Orchestration Controls] | CODE |
| MEDIUM | packages/core/README.md | 170 | | **Mix providers in one team** | 13 built-in providers plus any OpenAI-compatible endpoint (Ollama, vLLM, LM Studio, Op | CODE |
| MEDIUM | packages/core/README.md | 259 | - **[CodingScaffold](https://github.com/JRS1986/CodingScaffold)** — Agentic-coding scaffold that lists OMA as an optiona | CODE |
| MEDIUM | packages/core/README_zh.md | 170 | | **同队混用 provider** | 13 家内置 provider，外加任意 OpenAI 兼容端点（Ollama、vLLM、LM Studio、OpenRouter、Groq），同队可自由混用。将 tool call 作为纯文本输 | CODE |
| MEDIUM | packages/core/README_zh.md | 259 | - **[CodingScaffold](https://github.com/JRS1986/CodingScaffold)** — agentic-coding 脚手架，把 OMA 列为可选编排后端，附带 `runTeam` 工作流模板 | CODE |
| MEDIUM | packages/core/examples/README.md | 47 | ## patterns — orchestration patterns | COMMENT |
| MEDIUM | packages/core/src/agent/agent.ts | 147 | // External backend: it runs its own agentic loop, so none of the | COMMENT |
| MEDIUM | packages/core/src/agent/runner.ts | 848 | // Main agentic loop — `while (true)` until end_turn or maxTurns | COMMENT |
| MEDIUM | packages/create-oma-app/README.md | 15 | ├── src/index.ts # the demo: one goal → multi-agent DAG → dashboard | CODE |
| MEDIUM | packages/create-oma-app/README.md | 17 | ├── package.json # one runtime dependency: @open-multi-agent/core | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | …core/tests/e2e/coordinator-dependency-widening.test.ts | 151 | // ── Bug cases: brief declarations must not be widened ────────────────── | COMMENT |
| MEDIUM | …core/tests/e2e/coordinator-dependency-widening.test.ts | 167 | // ── Known-good: exhaustive declarations must still be fully wired ────── | COMMENT |
| MEDIUM | packages/core/examples/providers/gemma4-local.ts | 102 | // ═══════════════════════════════════════════════════════════════════════════ | COMMENT |
| MEDIUM | packages/core/examples/providers/gemma4-local.ts | 104 | // ═══════════════════════════════════════════════════════════════════════════ | COMMENT |
| MEDIUM | packages/core/examples/providers/gemma4-local.ts | 155 | // ═══════════════════════════════════════════════════════════════════════════ | COMMENT |
| MEDIUM | packages/core/examples/providers/gemma4-local.ts | 157 | // ═══════════════════════════════════════════════════════════════════════════ | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | packages/core/tests/short-circuit.test.ts | 201 | // | COMMENT |
| LOW | packages/core/examples/basics/multi-model-team.ts | 101 | // --------------------------------------------------------------------------- | COMMENT |
| LOW | …ges/core/examples/patterns/cross-provider-reasoning.ts | 21 | * Prerequisites: | COMMENT |
| LOW | …e/examples/cookbook/rare-disease-information-triage.ts | 501 | // Production source adapter sketch | COMMENT |
| LOW | …examples/cookbook/narrative-puzzle-hint-arbitration.ts | 561 | // Lore: load from game CMS with versioned narrative content | COMMENT |
| LOW | …kages/core/examples/cookbook/competitive-monitoring.ts | 441 | // | COMMENT |
| LOW | …examples/integrations/with-vercel-ai-sdk/next-env.d.ts | 1 | /// <reference types="next" /> | COMMENT |
| LOW | packages/core/src/llm/deepseek.ts | 21 | * by DeepSeek on 2026-07-24. | COMMENT |
| LOW | packages/core/src/llm/ai-sdk.ts | 221 | readonly name = 'ai-sdk' | COMMENT |
| LOW | packages/core/src/llm/hunyuan.ts | 41 | // continues a tool-calling conversation, otherwise the chain of thought is | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | …/examples/integrations/with-tencentdb-memory/README.md | 120 | -e MODEL_API_KEY="your-api-key" \ | CODE |