`mts1b-operations` — public API surface

Compliance + audit chain + halt manager + 10 watchdogs.

Halt manager

from mts1b_operations.halt import HaltManager

mgr = HaltManager()

# Request a halt (auto from watchdog, or manual)
req = HaltRequest(
    severity="FUND_HALT",
    fund_id="paper-momentum",
    reason="daily PL -3.2% breached -3.0%",
    requested_by="mts1b-riskengine",
)
await mgr.request(req)
# Published to mts.v1.operations.halt.requested

# Resume (sticky — needs operator)
await mgr.resume(fund_id="paper-momentum", cosigned_by="operator:mondipsen")

# Status
halts = await mgr.list_active()
status = await mgr.status(fund_id="paper-momentum")

CLI:

mts cmd halt                       # firm-wide; type HALT to confirm
mts cmd cancel-all                 # cancel every open order; type CANCEL
mts cmd flatten-paper               # flatten paper funds
mts cmd resume                      # firm-wide resume
mts cmd resume <fund_id>
mts cmd resume <strategy_id>
mts cmd restart-all                 # restart 5 mts1b-* services

Audit chain

from mts1b_operations.audit import audit_chain

# Append
entry = await audit_chain.append(
    actor="mts1b-oms",
    action="order_rejected",
    subject_id=order.order_id,
    data={"reason": "drawdown_halt", "envelope_id": "env-001"},
)
# AuditEntry(sequence=47832, hash="...", timestamp=...)

# Query
trail = await audit_chain.get_trail(subject_id="ord-abc")
# list[AuditEntry] in chronological order

# Verify integrity
ok = await audit_chain.verify(from_sequence=0)
# True or raises ChainIntegrityError

CLI:

mts mts1b-operations audit show --subject-id <order_id>
mts mts1b-operations audit verify --from-sequence 0
mts mts1b-operations audit tail

Watchdogs

10 background runners, each emits NATS events on alert:

Watchdog	Subject	Triggers when
`predictive_health`	`mts.v1.operations.watchdog.health`	NAV outside 99% CI
`drift_monitor`	`mts.v1.operations.watchdog.drift`	strategy IC drift
`vpin`	`mts.v1.operations.watchdog.vpin`	VPIN > 0.4
`news_spike`	`mts.v1.operations.watchdog.news`	news count > 5σ
`position_anomaly`	`mts.v1.operations.watchdog.position`	unusual position delta
`db_watchdog`	`mts.v1.operations.watchdog.db`	row count mismatch / slow query
`strategy_watchdog`	`mts.v1.operations.watchdog.strategy`	per-strategy PL outside CI
`theta_watchdog`	`mts.v1.operations.watchdog.theta`	net theta exposure breach
`dependency_watchdog`	`mts.v1.operations.watchdog.dependency`	upstream /healthz red 3x
`portfolio_vol_dd`	`mts.v1.operations.watchdog.vol_dd`	realized vol > 1.5× target

Each watchdog has a CLI subcommand:

mts mts1b-operations watchdog status
mts mts1b-operations watchdog config --watchdog vpin --threshold 0.4
mts mts1b-operations watchdog pause --watchdog news_spike

Market-data API (public read-only)

GET /v1/quotes/AAPL          → Quote
GET /v1/bars/AAPL?interval=1d → list[Bar]
GET /v1/funds                 → list[FundStatus]
GET /v1/halts                 → list[HaltRequest]
GET /v1/healthz               → {"status": "ok"}

Rate-limited per IP. No auth needed for top-level read.

Compliance / reg reporting (Wave 3)

from mts1b_operations.compliance.reg_reporting import cat

await cat.report_event(
    event_type="newOrderEvent",
    order=order,
    venue=venue,
    actor=actor,
)

Hookable adapters for CAT (Consolidated Audit Trail), OATS (legacy), MIFID-II. Off by default; enable per-fund:

fund:
  fund_id: live-equities
  compliance:
    cat_enabled: true
    cat_subscriber_id: "ABCD12345"

NATS subjects

Subject	Direction
`mts.v1.>` (everything)	subscribe
`mts.v1.operations.audit.appended`	publish
`mts.v1.operations.halt.requested`	publish
`mts.v1.operations.watchdog.*`	publish

Halt manager​

Audit chain​

Watchdogs​

Market-data API (public read-only)​

Compliance / reg reporting (Wave 3)​

NATS subjects​

See also​