# Send Doubleword API to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete. ``` ### Upgrade existing ```text I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "doubleword", "name": "Doubleword API", "source": "tencent", "type": "skill", "category": "开发工具", "sourceUrl": "https://clawhub.ai/pjb157/doubleword", "canonicalUrl": "https://clawhub.ai/pjb157/doubleword", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/doubleword", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=doubleword", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "SKILL.md", "scripts/create_batch_file.py", "references/pricing.md", "references/api_reference.md", "references/getting_started.md" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "slug": "doubleword", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-05-02T17:02:56.451Z", "expiresAt": "2026-05-09T17:02:56.451Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=doubleword", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=doubleword", "contentDisposition": "attachment; filename=\"doubleword-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "doubleword" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/doubleword" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/doubleword", "downloadUrl": "https://openagent3.xyz/downloads/doubleword", "agentUrl": "https://openagent3.xyz/skills/doubleword/agent", "manifestUrl": "https://openagent3.xyz/skills/doubleword/agent.json", "briefUrl": "https://openagent3.xyz/skills/doubleword/agent.md" } } ``` ## Documentation ### Doubleword Batch Inference Process multiple AI inference requests asynchronously using the Doubleword batch API with high throughput and low cost. ### Prerequisites Before submitting batches, you need: Doubleword Account - Sign up at https://app.doubleword.ai/ API Key - Create one in the API Keys section of your dashboard Account Credits - Add credits to process requests (see pricing below) ### When to Use Batches Batches are ideal for: Multiple independent requests that can run simultaneously Workloads that don't require immediate responses Large volumes that would exceed rate limits if sent individually Cost-sensitive workloads (24h window = 50-60% cheaper than realtime) Tool calling and structured output generation at scale ### Available Models & Pricing Pricing is per 1 million tokens (input / output): Qwen3-VL-30B-A3B-Instruct-FP8 (mid-size): Realtime SLA: $0.16 / $0.80 1-hour SLA: $0.07 / $0.30 (56% cheaper) 24-hour SLA: $0.05 / $0.20 (69% cheaper) Qwen3-VL-235B-A22B-Instruct-FP8 (flagship): Realtime SLA: $0.60 / $1.20 1-hour SLA: $0.15 / $0.55 (75% cheaper) 24-hour SLA: $0.10 / $0.40 (83% cheaper) Supports up to 262K total tokens, 16K new tokens per request Cost estimation: Upload files to the Doubleword Console to preview expenses before submitting. ### Quick Start Two ways to submit batches: Via API: Create JSONL file with requests Upload file to get file ID Create batch using file ID Poll status until complete Download results from output_file_id Via Web Console: Navigate to Batches section at https://app.doubleword.ai/ Upload JSONL file Configure batch settings (model, completion window) Monitor progress in real-time dashboard Download results when ready ### Step 1: Create Batch Request File Create a .jsonl file where each line contains a complete, valid JSON object with no line breaks within the object: {"custom_id": "req-1", "method": "POST", "url": "/v1/chat/completions", "body": {"model": "anthropic/claude-3-5-sonnet", "messages": [{"role": "user", "content": "What is 2+2?"}]}} {"custom_id": "req-2", "method": "POST", "url": "/v1/chat/completions", "body": {"model": "anthropic/claude-3-5-sonnet", "messages": [{"role": "user", "content": "What is the capital of France?"}]}} Required fields per line: custom_id: Unique identifier (max 64 chars) - use descriptive IDs like "user-123-question-5" for easier result mapping method: Always "POST" url: API endpoint - "/v1/chat/completions" or "/v1/embeddings" body: Standard API request with model and messages Optional body parameters: temperature: 0-2 (default: 1.0) max_tokens: Maximum response tokens top_p: Nucleus sampling parameter stop: Stop sequences tools: Tool definitions for tool calling (see Tool Calling section) response_format: JSON schema for structured outputs (see Structured Outputs section) File requirements: Max size: 200MB Format: JSONL only (JSON Lines - newline-delimited JSON) Each line must be valid JSON with no internal line breaks No duplicate custom_id values Split large batches into multiple files if needed Common pitfalls: Line breaks within JSON objects (will cause parsing errors) Invalid JSON syntax Duplicate custom_id values Helper script: Use scripts/create_batch_file.py to generate JSONL files programmatically: python scripts/create_batch_file.py output.jsonl Modify the script's requests list to generate your specific batch requests. ### Step 2: Upload File Via API: curl https://api.doubleword.ai/v1/files \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" \\ -F purpose="batch" \\ -F file="@batch_requests.jsonl" Via Console: Upload through the Batches section at https://app.doubleword.ai/ Response contains id field - save this file ID for next step. ### Step 3: Create Batch Via API: curl https://api.doubleword.ai/v1/batches \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" \\ -H "Content-Type: application/json" \\ -d '{ "input_file_id": "file-abc123", "endpoint": "/v1/chat/completions", "completion_window": "24h" }' Via Console: Configure batch settings in the web interface. Parameters: input_file_id: File ID from upload step endpoint: API endpoint ("/v1/chat/completions" or "/v1/embeddings") completion_window: Choose based on urgency and budget: "24h": Best pricing, results within 24 hours (typically faster) "1h": 50% price premium, results within 1 hour (typically faster) Realtime: Limited capacity, highest cost (batch service optimized for async) Response contains batch id - save this for status polling. Before submitting, verify: You have access to the specified model Your API key is active You have sufficient account credits ### Step 4: Poll Status Via API: curl https://api.doubleword.ai/v1/batches/batch-xyz789 \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" Via Console: Monitor real-time progress in the Batches dashboard. Status progression: validating - Checking input file format in_progress - Processing requests completed - All requests finished Other statuses: failed - Batch failed (check error_file_id) expired - Batch timed out cancelling/cancelled - Batch cancelled Response includes: output_file_id - Download results here error_file_id - Failed requests (if any) request_counts - Total/completed/failed counts Polling frequency: Check every 30-60 seconds during processing. Early access: Results available via output_file_id before batch fully completes - check X-Incomplete header. ### Step 5: Download Results Via API: curl https://api.doubleword.ai/v1/files/file-output123/content \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" \\ > results.jsonl Via Console: Download results directly from the Batches dashboard. Response headers: X-Incomplete: true - Batch still processing, more results coming X-Last-Line: 45 - Resume point for partial downloads Output format (each line): { "id": "batch-req-abc", "custom_id": "request-1", "response": { "status_code": 200, "body": { "id": "chatcmpl-xyz", "choices": [{ "message": { "role": "assistant", "content": "The answer is 4." } }] } } } Download errors (if any): curl https://api.doubleword.ai/v1/files/file-error123/content \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" \\ > errors.jsonl Error format (each line): { "id": "batch-req-def", "custom_id": "request-2", "error": { "code": "invalid_request", "message": "Missing required parameter" } } ### Tool Calling in Batches Tool calling (function calling) enables models to intelligently select and use external tools. Doubleword maintains full OpenAI compatibility. Example batch request with tools: { "custom_id": "tool-req-1", "method": "POST", "url": "/v1/chat/completions", "body": { "model": "anthropic/claude-3-5-sonnet", "messages": [{"role": "user", "content": "What's the weather in Paris?"}], "tools": [{ "type": "function", "function": { "name": "get_weather", "description": "Get current weather for a location", "parameters": { "type": "object", "properties": { "location": {"type": "string"} }, "required": ["location"] } } }] } } Use cases: Agents that interact with APIs at scale Fetching real-time information for multiple queries Executing actions through standardized tool definitions ### Structured Outputs in Batches Structured outputs guarantee that model responses conform to your JSON Schema, eliminating issues with missing fields or invalid enum values. Example batch request with structured output: { "custom_id": "structured-req-1", "method": "POST", "url": "/v1/chat/completions", "body": { "model": "anthropic/claude-3-5-sonnet", "messages": [{"role": "user", "content": "Extract key info from: John Doe, 30 years old, lives in NYC"}], "response_format": { "type": "json_schema", "json_schema": { "name": "person_info", "schema": { "type": "object", "properties": { "name": {"type": "string"}, "age": {"type": "integer"}, "city": {"type": "string"} }, "required": ["name", "age", "city"] } } } } } Benefits: Guaranteed schema compliance No missing required keys No hallucinated enum values Seamless OpenAI compatibility ### autobatcher: Automatic Batching autobatcher is a Python client that automatically converts individual API calls into batched requests, reducing costs without code changes. Installation: pip install autobatcher How it works: Collection Phase: Requests accumulate during a time window (default: 1 second) or until batch size threshold Batch Submission: Collected requests are submitted together Result Polling: System monitors for completed responses Transparent Response: Your code receives standard ChatCompletion responses Key benefit: Significant cost reduction through automatic batching while writing normal async code using the familiar OpenAI interface. Documentation: https://github.com/doublewordai/autobatcher ### List All Batches Via API: curl https://api.doubleword.ai/v1/batches?limit=10 \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" Via Console: View all batches in the dashboard. ### Cancel Batch Via API: curl https://api.doubleword.ai/v1/batches/batch-xyz789/cancel \\ -X POST \\ -H "Authorization: Bearer $DOUBLEWORD_API_KEY" Via Console: Click cancel in the batch details view. Notes: Unprocessed requests are cancelled Already-processed results remain downloadable Only charged for completed work Cannot cancel completed batches ### Processing Results Parse JSONL output line-by-line: import json with open('results.jsonl') as f: for line in f: result = json.loads(line) custom_id = result['custom_id'] content = result['response']['body']['choices'][0]['message']['content'] print(f"{custom_id}: {content}") ### Handling Partial Results Check for incomplete batches and resume: import requests response = requests.get( 'https://api.doubleword.ai/v1/files/file-output123/content', headers={'Authorization': f'Bearer {api_key}'} ) if response.headers.get('X-Incomplete') == 'true': last_line = int(response.headers.get('X-Last-Line', 0)) print(f"Batch incomplete. Processed {last_line} requests so far.") # Continue polling and download again later ### Retry Failed Requests Extract failed requests from error file and resubmit: import json failed_ids = [] with open('errors.jsonl') as f: for line in f: error = json.loads(line) failed_ids.append(error['custom_id']) print(f"Failed requests: {failed_ids}") # Create new batch with only failed requests ### Processing Tool Calls Handle tool call responses: import json with open('results.jsonl') as f: for line in f: result = json.loads(line) message = result['response']['body']['choices'][0]['message'] if message.get('tool_calls'): for tool_call in message['tool_calls']: print(f"Tool: {tool_call['function']['name']}") print(f"Args: {tool_call['function']['arguments']}") ### Best Practices Descriptive custom_ids: Include context in IDs for easier result mapping Good: "user-123-question-5", "dataset-A-row-42" Bad: "1", "req1" Validate JSONL locally: Ensure each line is valid JSON with no internal line breaks before upload No duplicate IDs: Each custom_id must be unique within the batch Split large files: Keep under 200MB limit by splitting into multiple batches Choose appropriate window: Use 24h for cost savings (50-83% cheaper), 1h only when time-sensitive Handle errors gracefully: Always check error_file_id and retry failed requests Monitor request_counts: Track progress via completed/total ratio Save file IDs: Store batch_id, input_file_id, output_file_id for later retrieval Use cost estimator: Preview expenses in console before submitting large batches Consider autobatcher: For ongoing workloads, use autobatcher to automatically batch individual API calls ### Reference Documentation For complete API details, see: API Reference: references/api_reference.md - Full endpoint documentation and schemas Getting Started Guide: references/getting_started.md - Detailed setup and account management Pricing Details: references/pricing.md - Model costs and SLA comparison ## Trust - Source: tencent - Verification: Indexed source record - Publisher: pjb157 - Version: 1.0.0 ## Source health - Status: healthy - Item download looks usable. - Yavira can redirect you to the upstream package for this item. - Health scope: item - Reason: direct_download_ok - Checked at: 2026-05-02T17:02:56.451Z - Expires at: 2026-05-09T17:02:56.451Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/doubleword) - [Send to Agent page](https://openagent3.xyz/skills/doubleword/agent) - [JSON manifest](https://openagent3.xyz/skills/doubleword/agent.json) - [Markdown brief](https://openagent3.xyz/skills/doubleword/agent.md) - [Download page](https://openagent3.xyz/downloads/doubleword)