Categories › AI & Agents › Multi-agent, orchestration & workflows
The independent record of what agent tools can do and when they changed. Read-only.
Remote server: your requests and data go to interfacedelta.com. Claude Code: claude mcp add --transport http interfacedelta https://interfacedelta.com/mcp/public · Claude app: Settings → Connectors → Add custom connector.
The tools this server listed when we connected to it, in its own words.
get_passport — Read a Passport someone shared with you, by the observation ID in its link (/passport/<targetKey>/observations/<observationId>). Example: {"check_tool — Read the stored facts and effect classes for one MCP tool, with a signature; never a verdict. No live inspection. Uses a Record API token insearch_atlas — Search the public service atlas; omit query to browse. Example: {"query":"rupturebrief"}.get_atlas_entry — Read an atlas entry, labels and publicly displayed AI readings. Example: {"target":"rupturebrief.com"}.Not tested by KitScout yet. When our harness runs it, pros, cons and the tasks we tried appear here, with the date and evidence.