research

Run multi-pass deep research across the current infrastructure estate.

clanker research fans out across analysis subagents, prioritizes cost and operational findings, and emits structured JSON that Clanker Cloud can render into a report.

Usage

clanker research [question] [flags]

Examples

clanker research "where are we wasting the most cloud spend?"
clanker research "find resilience risks across prod" --profile prod
clanker research "compare app runtime options for this estate" --gcp-project my-project
clanker research "summarize open infra risks" --workspace production

Provider Scope Flags

Flag	Description
--profile	AWS profile for provider-side research helpers
--gcp-project	GCP project ID for provider-side research helpers
--azure-subscription	Azure subscription ID for provider-side research helpers
--workspace	Terraform workspace for provider-side research helpers

AI Override Flags

research accepts the same AI profile, key, and model override flags as ask, including --ai-profile, --openai-key, --openai-model, --anthropic-key, --gemini-key, --cohere-key, --deepseek-key, --minimax-key, and --local-model-inference-url.

Output

The command is intended for report generation and agent handoff. Prefer --debug while validating a new workspace so you can see provider-helper failures and missing credentials.