160-tool MCP server for real-time global intelligence from free public APIs: markets, liquidity, prediction markets, SEC filings, conflict, military, aircraft-cohort early warning, migration and border flows, cyber, climate, news and more. Dated observation history, 29 research prompts, cited briefs, geofences, SSE dashboard. MIT, no paid API keys.
Python ai-toolsai-watchanthropicclaudecybersecuritydashboardgeofencinggeopoliticalgeospatialintelligencemarketsmcpmigrationmilitarymodel-context-protocolnuclearosintreal-timespace-weatherthreat-intelligence
"""Domain-modular tool registry (Phase 26 server.py split).
Each domain module in this package exports two names:
- ``TOOLS: list[mcp.types.Tool]`` — the domain's tool definitions.
- ``HANDLERS: dict[str, Handler]`` — tool name to handler, where a
handler is ``async def h(arguments: dict) -> Any`` and reaches shared
infrastructure through ``world_intel_mcp.runtime`` (fetcher, cache,
aoi_store, vector_store).
``aggregate()`` builds the combined registry and enforces, at import
time, the invariant ROADMAP.md has carried since Phase 0: every
registered tool has exactly one handler and every handler has exactly
one registration, with no name collisions across modules. A drift that
the old text-scan parity check could only catch in CI now prevents the
server from importing at all.
"""
from typing import Any, Awaitable, Callable
from mcp.types import Tool
Handler = Callable[[dict[str, Any]], Awaitable[Any]]
from . import ( # noqa: E402
aoi,
conflict,
day_timeline,
finance,
geospatial,
hazards,
history,
history_backfill,
infrastructure,
intelligence,
lagged_connections,
liquidity,
markets,
news_market,
news_signals,
prediction_markets,
society,
synthesis,
system,
vector,
)
# Module order defines list_tools order, approximating the pre-split
# registry order (markets first ... system last).
_MODULES = [
markets,
hazards,
conflict,
infrastructure,
society,
prediction_markets,
intelligence,
geospatial,
synthesis,
news_signals,
finance,
liquidity,
news_market,
vector,
aoi,
history,
lagged_connections,
day_timeline,
history_backfill,
system,
]
def aggregate(
modules: list,
) -> tuple[list[Tool], dict[str, Handler]]:
"""Combine per-domain registries, refusing drift and collisions."""
all_tools: list[Tool] = []
all_handlers: dict[str, Handler] = {}
for module in modules:
tools: list[Tool] = module.TOOLS
handlers: dict[str, Handler] = module.HANDLERS
tool_names = {t.name for t in tools}
if tool_names != set(handlers):
raise RuntimeError(
f"{module.__name__}: TOOLS/HANDLERS drift - "
f"registered-without-handler {sorted(tool_names - set(handlers))}, "
f"handler-without-registration {sorted(set(handlers) - tool_names)}"
)
collisions = tool_names & set(all_handlers)
if collisions:
raise RuntimeError(
f"{module.__name__}: tool name collision across modules: "
f"{sorted(collisions)}"
)
all_tools.extend(tools)
all_handlers.update(handlers)
return all_tools, all_handlers
ALL_TOOLS, ALL_HANDLERS = aggregate(_MODULES)