juspay/NeuroLink AI

Run AI-powered workflows with 13 providers (OpenAI, Anthropic, Google, AWS, Azure, and more)

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Stale
License
None
namedescriptionrequireddefault
promptThe prompt to send to the AI modelyes
openai_api_keyOpenAI API keyno
anthropic_api_keyAnthropic API keyno
google_ai_api_keyGoogle AI Studio API keyno
azure_openai_api_keyAzure OpenAI API keyno
azure_openai_endpointAzure OpenAI endpoint URLno
azure_openai_deploymentAzure OpenAI deployment nameno
mistral_api_keyMistral AI API keyno
huggingface_api_keyHugging Face API keyno
openrouter_api_keyOpenRouter API keyno
litellm_api_keyLiteLLM API keyno
litellm_base_urlLiteLLM base URLno
openai_compatible_api_keyOpenAI-compatible API keyno
openai_compatible_base_urlOpenAI-compatible base URLno
aws_access_key_idAWS Access Key ID for Bedrock/SageMakerno
aws_secret_access_keyAWS Secret Access Keyno
aws_regionAWS Regionnous-east-1
aws_session_tokenAWS Session Tokenno
bedrock_model_idAWS Bedrock model IDno
sagemaker_endpointAmazon SageMaker endpointno
google_vertex_projectGoogle Cloud project ID for Vertex AIno
google_vertex_locationGoogle Cloud locationnous-central1
google_application_credentialsGCP service account JSON (base64 encoded)no
providerAI provider (openai, anthropic, google-ai, vertex, azure, bedrock, mistral, huggingface, openrouter, litellm, ollama, sagemaker, openai-compatible)noauto
modelSpecific model to useno
temperatureSampling temperature (0.0-2.0)no0.7
max_tokensMaximum tokens in responseno4096
system_promptSystem prompt for contextno
commandCLI command: generate, stream, batchnogenerate
image_pathsComma-separated image pathsno
pdf_pathsComma-separated PDF pathsno
csv_pathsComma-separated CSV pathsno
video_pathsComma-separated video pathsno
thinking_enabledEnable extended thinkingnofalse
thinking_levelThinking level: minimal, low, medium, highnomedium
thinking_budgetThinking token budgetno10000
enable_analyticsEnable usage analytics and cost trackingnofalse
enable_evaluationEnable response quality evaluationnofalse
output_formatOutput format: text, jsonnotext
output_fileOutput file pathno
enable_toolsEnable MCP toolsnofalse
mcp_config_pathPath to .mcp-config.json fileno
post_commentPost AI response as PR/issue commentnofalse
update_existing_commentUpdate existing NeuroLink comment instead of creating newnotrue
comment_tagHTML comment tag to identify NeuroLink commentsnoneurolink-action
github_tokenGitHub token for PR/issue operationsno${{ github.token }}
timeoutRequest timeout in secondsno300
debugEnable debug loggingnofalse
neurolink_versionNeuroLink CLI version to installnolatest
working_directoryWorking directory for CLI executionno.
namedescription
responseAI response text content
response_jsonFull JSON response including metadata
providerProvider that was used
modelModel that was used
tokens_usedTotal tokens consumed
prompt_tokensInput/prompt tokens
completion_tokensOutput/completion tokens
costEstimated cost in USD (if analytics enabled)
execution_timeExecution time in milliseconds
evaluation_scoreQuality evaluation score 0-100 (if evaluation enabled)
comment_idGitHub comment ID (if post_comment enabled)
errorError message if execution failed