> ## Documentation Index > Fetch the complete documentation index at: https://docs.mka1.com/llms.txt > Use this file to discover all available pages before exploring further. # Import a finished coding-agent run > Uploads one completed harness job — the run and its trials — directly, so results never round-trip through external storage the way the Hugging Face importer does. Every `task_id` must already exist in the target suite version's manifest; unknown ids are rejected rather than invented, because a typo would otherwise import as a task the suite does not have and read as a model that never covered the version. The run is recorded as `completed` and is not executed: this is a write, not a run request, so no workflow is started. Trials that cover every task of the version get `task_ids: null`, which is what lets the run rank as a complete sweep. Trajectories are NOT part of this payload — a full 91-task run measured 8.3 MB of them — so upload each one against `PUT /evals/runs/{run_id}/samples/{sample_index}/trajectory` afterwards. Pass `idempotency_key` and a retried upload returns the first run instead of doubling the trials behind a board row. ## OpenAPI ````yaml https://apigw.mka1.com/speakeasy.json post /api/v1/llm/evals/imports/agent-run openapi: 3.1.1 info: title: MKA1 API version: 1.1.0 description: >- The MKA1 API is a RESTful API that provides access to the MKA1 platform. Learn how to get started with the API and the TypeScript SDK [here](https://mka1.apidocumentation.com/guides/getting-started). license: name: Proprietary servers: - url: https://apigw.mka1.com description: MKA1 API Gateway - url: / description: Relative server URL (configurable via SDK constructor) security: [] tags: - name: Resource Authorization description: >- Manage permissions for LLM resources. Create resources, grant/revoke permissions, and delete resources. Only resource owners can grant, revoke, or delete permissions. x-displayName: Resource Authorization - name: Embeddings description: >- Text embedding API endpoints for generating vector representations of text. Create semantic embeddings for search, clustering, and similarity matching using various embedding models. x-displayName: Embeddings - name: Feedback description: >- User feedback API for rating and commenting on chat completions. Collect thumbs up/down ratings and detailed feedback to improve model responses and track user satisfaction. x-displayName: Feedback - name: Images description: >- Image generation API endpoints for creating images from text descriptions. Generate images with control over size, quality, and style. x-displayName: Images - name: MCP Vault description: >- MCP vault API for storing user-owned MCP server configurations and encrypted credentials. Agents reference vault IDs so secrets are resolved only at tool execution time. x-displayName: MCP Vault - name: Speech description: >- Speech API endpoints for audio processing. Convert text to natural-sounding speech (TTS) or transcribe speech to text (STT) in different languages. x-displayName: Speech - name: Usage description: >- Usage tracking and analytics API for monitoring token consumption, request counts, and cost analysis. View detailed statistics per user, model, and time period. x-displayName: Usage - name: Extract description: >- Structured data extraction API for extracting information from files. Define JSON schemas to extract structured data from images, PDFs, and documents. Supports reusable schema templates. x-displayName: Extract - name: Text Classification description: >- Text classification API for categorizing text into predefined labels. Use AI models to classify text content for sentiment analysis, topic categorization, and content moderation. x-displayName: Text Classification - name: Responses description: >- Agent-powered responses API for creating AI agents with autonomous tool usage. Build conversational assistants that can use web search, file operations, image generation, code execution, computer use simulation, and MCP integrations. Supports background processing, streaming, and real-time status tracking. x-displayName: Responses - name: Files description: >- File management API for uploading, storing, and managing files with automatic expiration and S3 integration. Upload files that can be used with Assistants, Vector Stores, and other features. Files are stored in S3 with metadata tracked in PostgreSQL. Supports automatic cleanup of expired files. x-displayName: Files - name: Vector Stores description: >- Vector store API for storing and searching documents using embeddings. Create vector stores, upload files with automatic chunking and embedding generation, and perform semantic search. Files are processed asynchronously using Temporal workflows for durability. Supports automatic cleanup of expired stores and LanceDB for efficient vector storage. x-displayName: Vector Stores - name: Conversations description: >- Conversation management API for storing and retrieving conversation state across Response API calls. Create conversations, add items (user messages, assistant messages, system messages), and maintain conversation history. Supports metadata tracking and multi-turn dialogue state management. x-displayName: Conversations - name: Guardrails description: >- AI safety guardrails API for configuring content moderation and security policies. Set up ban word lists, prompt injection detection, and system prompt leakage prevention. Guardrails apply to all requests from an account and can be tested before deployment. x-displayName: Guardrails - name: Models description: >- Model listing API for discovering available models. Returns model IDs, ownership, and metadata for all registered models in the gateway. x-displayName: Models - name: Skills description: >- Skills API for managing versioned bundles of instructions and files following the Agent Skills standard. Create, version, and download reusable skill packages that include SKILL.md manifests for agent environments. x-displayName: Skills - name: Chat Completions description: >- **Deprecated: Use the Responses API (`/api/v1/llm/responses`) instead.** Chat completion endpoints with support for streaming, tool calls, and multiple providers. x-deprecated: true x-displayName: Chat Completions - name: Batches x-displayName: Batches - name: Evals x-displayName: Evals - name: Fine-Tuning x-displayName: Fine-Tuning - name: Memory Stores x-displayName: Memory Stores - name: Prompts x-displayName: Prompts - name: API Key x-displayName: API Key - name: Session x-displayName: Session - name: Organization x-displayName: Organization - name: Cluster Admin x-displayName: Cluster Admin - name: Sessions description: Create, inspect, access, and terminate sandbox sessions. x-displayName: Sessions - name: Browser description: >- Connect to browser sessions through the gateway port proxy. Browser sessions expose a Chrome DevTools Protocol endpoint on port 9222. x-displayName: Browser - name: Execution description: Run shell commands and code inside an existing sandbox session. x-displayName: Execution - name: Workspace description: >- Inspect the workspace manifest, transfer files or archives, and download generated artifacts. x-displayName: Workspace - name: Sandbox Usage description: >- Aggregate sandbox usage statistics across sessions, execution, and workspace operations. x-displayName: Sandbox Usage - name: Sandbox Pricing description: >- Cluster-admin management of the sandbox compute rate card used for budgeted spend. x-displayName: Sandbox Pricing - name: schema-3_other x-displayName: other - name: Agents description: Create and manage reusable agent definitions. x-displayName: Agents - name: Agent Versions description: Inspect an agent's configuration history and roll back to a prior version. x-displayName: Agent Versions - name: Agent Runs description: Execute saved agents and inspect persisted run results. x-displayName: Agent Runs - name: Agent Connectors description: >- Connect saved agents to external messaging channels such as Telegram, including text, photo, and supported document exchange. x-displayName: Agent Connectors - name: Agent Schedules description: Create and manage scheduled or recurring saved agent runs. x-displayName: Agent Schedules - name: schema-4_other x-displayName: other - name: Budgets x-displayName: Budgets - name: Settings x-displayName: Settings - name: Compute Jobs x-displayName: Compute Jobs - name: Compute Services x-displayName: Compute Services - name: Compute Catalog x-displayName: Compute Catalog - name: Compute Volumes x-displayName: Compute Volumes - name: Compute Pricing x-displayName: Compute Pricing - name: Compute Secrets x-displayName: Compute Secrets - name: Compute Usage x-displayName: Compute Usage - name: Compute Tenants x-displayName: Compute Tenants - description: Create, list, read, update, and delete repositories. name: Repositories x-displayName: Repositories paths: /api/v1/llm/evals/imports/agent-run: post: tags: - Evals summary: Import a finished coding-agent run description: >- Uploads one completed harness job — the run and its trials — directly, so results never round-trip through external storage the way the Hugging Face importer does. Every `task_id` must already exist in the target suite version's manifest; unknown ids are rejected rather than invented, because a typo would otherwise import as a task the suite does not have and read as a model that never covered the version. The run is recorded as `completed` and is not executed: this is a write, not a run request, so no workflow is started. Trials that cover every task of the version get `task_ids: null`, which is what lets the run rank as a complete sweep. Trajectories are NOT part of this payload — a full 91-task run measured 8.3 MB of them — so upload each one against `PUT /evals/runs/{run_id}/samples/{sample_index}/trajectory` afterwards. Pass `idempotency_key` and a retried upload returns the first run instead of doubling the trials behind a board row. operationId: importAgentEvalRun parameters: - name: X-On-Behalf-Of in: header required: false schema: type: string description: Optional external end-user identifier forwarded by the API gateway. requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/ImportAgentRunRequest' example: suite_id: eval_suite_aa87e2b1112a455b8deabed784372198 suite_version: 1 agent_name: omp agent_version: 17.3.8 agent_effort: medium model: meetkai:functionary-urdu-mini-pak cost_usd: 3.77 created_at: 1704067200 idempotency_key: harbor-job-2026-08-23-luna-medium metadata: harness: harbor benchmark: terminal-bench-2.1 trials: - task_id: fix-permissions sample_index: 0 status: completed prompt: >- The test suite fails because the fixture directory is not writable. Fix it. target: all tests pass output_text: 9 passed in 2.25s scores: reward: 1 resolved: 1 cost_usd: 0.0279 input_tokens: 658456 output_tokens: 6489 cached_tokens: 658366 responses: '200': description: OK content: application/json: schema: type: object properties: id: type: string object: const: eval.run suite_id: type: string suite_version: type: integer minimum: -9007199254740991 maximum: 9007199254740991 suite_version_id: type: string org_id: type: string description: The org that owns this run. team_id: type: string description: The team that owns this run. status: $ref: '#/components/schemas/EvalRunStatus' models: type: array items: type: string task_ids: anyOf: - type: array items: type: string - type: 'null' judge_model: anyOf: - type: string - type: 'null' embedding_model: anyOf: - type: string - type: 'null' generation: type: object properties: instructions: type: string temperature: type: number minimum: 0 maximum: 2 top_p: type: number minimum: 0 maximum: 1 max_output_tokens: type: integer minimum: 1 maximum: 9007199254740991 max_gen_toks: type: integer minimum: 1 maximum: 9007199254740991 description: lm-eval alias for max_output_tokens. stop: anyOf: - type: string - type: array minItems: 1 items: type: string until: type: array minItems: 1 items: type: string description: lm-eval generate_until stop sequences. max_tool_calls: type: integer minimum: 1 maximum: 9007199254740991 reasoning: {} text: {} tools: type: array items: anyOf: - {} - type: 'null' tool_choice: {} parallel_tool_calls: type: boolean truncation: enum: - auto - disabled type: string service_tier: enum: - auto - default - flex - priority type: string presence_penalty: type: number minimum: -2 maximum: 2 frequency_penalty: type: number minimum: -2 maximum: 2 top_k: type: integer minimum: 0 maximum: 9007199254740991 min_p: type: number minimum: 0 maximum: 1 repetition_penalty: type: number minimum: 0 do_sample: type: boolean extra_body: type: object propertyNames: type: string additionalProperties: {} chat_template_kwargs: type: object propertyNames: type: string additionalProperties: {} prefill_think: anyOf: - type: boolean - type: string use_cache: type: boolean timeout_seconds: type: integer minimum: 1 maximum: 3600 max_retries: type: integer minimum: 0 maximum: 10 max_empty_retries: type: integer minimum: 0 maximum: 10 additionalProperties: {} request_counts: type: object properties: total: type: integer minimum: -9007199254740991 maximum: 9007199254740991 completed: type: integer minimum: -9007199254740991 maximum: 9007199254740991 failed: type: integer minimum: -9007199254740991 maximum: 9007199254740991 required: - total - completed - failed metrics: anyOf: - type: object propertyNames: type: string additionalProperties: {} - type: 'null' error: anyOf: - type: object propertyNames: type: string additionalProperties: {} - type: 'null' artifact_file_ids: type: array items: type: string metadata: type: object propertyNames: type: string maxLength: 64 additionalProperties: type: string maxLength: 512 agent_name: anyOf: - type: string - type: 'null' description: >- Coding-agent harness that produced this run (e.g. `omp`, `claude-code`). Null for every non-harness eval kind — a null here is not a missing value, it means the run is not an agent run. agent_version: anyOf: - type: string - type: 'null' description: >- Version of `agent_name`, as the harness reported it. Null when `agent_name` is. agent_effort: anyOf: - type: string - type: 'null' description: >- Reasoning-effort setting the agent ran at (e.g. `medium`, `xhigh`). Part of the leaderboard's row identity: the same agent and model at two efforts are two rows, not one. Null when `agent_name` is. cost_usd: anyOf: - type: number - type: 'null' description: >- Total spend for the run in USD. Null when the harness did not report cost. created_at: type: integer minimum: -9007199254740991 maximum: 9007199254740991 started_at: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' completed_at: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' cancelled_at: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' failed_at: anyOf: - type: integer minimum: -9007199254740991 maximum: 9007199254740991 - type: 'null' required: - id - object - suite_id - suite_version - suite_version_id - org_id - team_id - status - models - task_ids - judge_model - embedding_model - generation - request_counts - metrics - error - artifact_file_ids - metadata - agent_name - agent_version - agent_effort - cost_usd - created_at - started_at - completed_at - cancelled_at - failed_at example: id: eval_run_aa87e2b1112a455b8deabed784372198 object: eval.run suite_id: eval_suite_aa87e2b1112a455b8deabed784372198 suite_version: 1 suite_version_id: eval_sver_aa87e2b1112a455b8deabed784372198 org_id: org-acme team_id: team-research status: completed models: - auto task_ids: null judge_model: auto embedding_model: auto generation: temperature: 0 max_output_tokens: 512 request_counts: total: 100 completed: 10 failed: 0 metrics: null error: null artifact_file_ids: [] metadata: purpose: mvp agent_name: omp agent_version: null agent_effort: null cost_usd: null created_at: 1704067200 started_at: 1704067210 completed_at: null cancelled_at: null failed_at: null security: - bearerAuth: [] x-codeSamples: - lang: python label: Python (SDK) source: |- from meetkai_mka1 import SDK with SDK( bearer_auth="", ) as sdk: res = sdk.llm.evals.import_agent_run(suite_id="eval_suite_aa87e2b1112a455b8deabed784372198", agent_name="omp", model="meetkai:functionary-urdu-mini-pak", suite_version=1, agent_version="17.3.8", agent_effort="medium", cost_usd=3.77, created_at=1704067200, metadata={ "harness": "harbor", "benchmark": "terminal-bench-2.1", }, idempotency_key="harbor-job-2026-08-23-luna-medium", trials=[ { "task_id": "fix-permissions", "sample_index": 0, "prompt": "The test suite fails because the fixture directory is not writable. Fix it.", "target": "all tests pass", "output_text": "9 passed in 2.25s", "scores": { "reward": 1, "resolved": 1, }, "cost_usd": 0.0279, "input_tokens": 658456, "output_tokens": 6489, "cached_tokens": 658366, }, ], finalize=True) # Handle response print(res) - lang: typescript label: Typescript (SDK) source: |- import { SDK } from "@meetkai/mka1"; const sdk = new SDK({ bearerAuth: "", }); async function run() { const result = await sdk.llm.evals.importAgentRun({ importAgentRunRequest: { suiteId: "eval_suite_aa87e2b1112a455b8deabed784372198", suiteVersion: 1, agentName: "omp", agentVersion: "17.3.8", agentEffort: "medium", model: "meetkai:functionary-urdu-mini-pak", costUsd: 3.77, createdAt: 1704067200, metadata: { "harness": "harbor", "benchmark": "terminal-bench-2.1", }, idempotencyKey: "harbor-job-2026-08-23-luna-medium", trials: [ { taskId: "fix-permissions", sampleIndex: 0, prompt: "The test suite fails because the fixture directory is not writable. Fix it.", target: "all tests pass", outputText: "9 passed in 2.25s", scores: { "reward": 1, "resolved": 1, }, costUsd: 0.0279, inputTokens: 658456, outputTokens: 6489, cachedTokens: 658366, }, ], }, }); console.log(result); } run(); - lang: csharp label: CSharp (SDK) source: >- using MeetKai.MKA1; using MeetKai.MKA1.Types.Components; using System.Collections.Generic; var sdk = new SDK(bearerAuth: ""); var res = await sdk.Llm.Evals.ImportAgentRunAsync(body: new ImportAgentRunRequest() { SuiteId = "eval_suite_aa87e2b1112a455b8deabed784372198", SuiteVersion = 1, AgentName = "omp", AgentVersion = "17.3.8", AgentEffort = "medium", Model = "meetkai:functionary-urdu-mini-pak", CostUsd = 3.77D, CreatedAt = 1704067200, Metadata = new Dictionary() { { "harness", "harbor" }, { "benchmark", "terminal-bench-2.1" }, }, IdempotencyKey = "harbor-job-2026-08-23-luna-medium", Trials = new List() { new ImportAgentRunRequestTrial() { TaskId = "fix-permissions", SampleIndex = 0, Prompt = "The test suite fails because the fixture directory is not writable. Fix it.", Target = "all tests pass", OutputText = "9 passed in 2.25s", Scores = new Dictionary() { { "reward", 1D }, { "resolved", 1D }, }, CostUsd = 0.0279D, InputTokens = 658456, OutputTokens = 6489, CachedTokens = 658366, }, }, }); // handle response components: schemas: ImportAgentRunRequest: type: object properties: suite_id: type: string minLength: 1 description: >- Existing suite to import into. Its manifest defines the task set, so create the suite first. suite_version: type: integer minimum: 1 maximum: 2147483647 description: Defaults to the suite's active version. agent_name: type: string minLength: 1 maxLength: 128 description: The harness agent, e.g. `omp`. agent_version: anyOf: - type: string maxLength: 64 - type: 'null' agent_effort: anyOf: - type: string maxLength: 32 - type: 'null' description: >- Reasoning-effort setting, e.g. `medium`. Part of the leaderboard's row identity, so the same agent and model at two efforts stay two rows. model: type: string minLength: 1 maxLength: 255 cost_usd: anyOf: - type: number minimum: 0 - type: 'null' description: >- Total spend for the job. Falls back to the sum of the trials' own costs when omitted. created_at: type: integer minimum: -9007199254740991 maximum: 9007199254740991 exclusiveMinimum: 0 description: >- When the job ran, unix seconds. Defaults to now — set it when backfilling, since the board orders by this. metadata: type: object propertyNames: type: string maxLength: 64 additionalProperties: type: string maxLength: 512 idempotency_key: type: string minLength: 1 maxLength: 255 description: >- Reusing a key returns the run created the first time instead of importing a second copy. Without one, a retried upload silently doubles the trials behind a board row. trials: type: array maxItems: 1000 items: type: object properties: task_id: type: string minLength: 1 maxLength: 255 description: >- Must already exist in the target suite version's manifest; unknown ids are rejected rather than invented. sample_index: type: integer minimum: 0 maximum: 2147483647 description: >- Defaults to the trial's position in `trials`. Unique per (task, model), not per run. status: enum: - completed - failed type: string default: completed description: >- `failed` means no score was produced at all. An agent timeout is NOT that — the harness still runs the verifier and scores the trial, so it stays `completed` and carries its exception in `error`. prompt: type: string default: '' description: The task instruction. target: anyOf: - type: string - type: 'null' output_text: anyOf: - type: string - type: 'null' description: >- What decided the score — the verifier's output. The agent's session goes to the trajectory route. scores: type: object propertyNames: type: string additionalProperties: type: number default: {} description: >- Metric values for this trial, e.g. `{ "reward": 1, "resolved": 1 }`. The board averages `reward` by default. error: anyOf: - type: object propertyNames: type: string additionalProperties: {} - type: 'null' description: >- The exception this trial recorded, if any. Does not by itself make the trial a failure. cost_usd: anyOf: - type: number minimum: 0 - type: 'null' input_tokens: anyOf: - type: integer minimum: 0 maximum: 9007199254740991 - type: 'null' output_tokens: anyOf: - type: integer minimum: 0 maximum: 9007199254740991 - type: 'null' cached_tokens: anyOf: - type: integer minimum: 0 maximum: 9007199254740991 - type: 'null' description: >- The share of `input_tokens` served from the provider's prompt cache. phase_timings: anyOf: - type: object propertyNames: type: string additionalProperties: {} - type: 'null' required: - task_id default: [] description: >- May be empty when opening a run to stream trials into. A one-shot import sends them all here. expected_trials: type: integer minimum: 0 maximum: 1000 description: >- How many trials the job will produce. Sets the run's total up front so the console shows real progress (12/89) while it streams, instead of 12/12 that grows. finalize: type: boolean default: true description: >- False opens the run `in_progress` for `POST /evals/runs/{run_id}/trials` to fill. Leaderboards only count completed runs, so a streaming run stays off the board until it is finalized. required: - suite_id - agent_name - model EvalRunStatus: enum: - queued - in_progress - finalizing - completed - failed - cancelling - cancelled type: string securitySchemes: bearerAuth: type: http scheme: bearer bearerFormat: API Key description: >- Gateway auth: send `Authorization: Bearer `. For multi-user server-side integrations, you can also send `X-On-Behalf-Of: `. ````