Intel

Anvil's research-agent loop. Trigger a scan; the agent looks for new LLMs, agent runners, IDEs, MCP servers, and capability shifts. Output is dual-purpose: Anvil's registries get proposed updates here, and a longer-form research note lands as an Anvil memory AND (optionally) as a markdown file under salud/research/portal/anvil-intel/.

Run a scan

Topic (optional — leave blank for everything)Lookback (days)

Save research note as an Anvil memory Write to salud/research/portal/anvil-intel/

Pending proposals (0)

No pending proposals. Run a scan above, or have a Claude session call anvil_propose_model / anvil_propose_tool.