Strategies

Vulcan’s strategy command group hosts long-running execution loops that are too stateful for one-shot orders. Each run is ledger-backed, supports pause / resume / finalize lifecycle commands, and can be started in detached mode so the runner keeps ticking in the background. Tick logs and ledgers are persisted under ~/.vulcan/strategy-runs. Three runners ship today:

TWAP — split a target size across timed slices.
Grid — maintain layered limit orders across a price band.
TA — rule-based runner over technical indicators for entry and exit.

Execution modes

Every strategy accepts --mode. Pick the one that matches your tolerance for live risk while testing:

Mode	Behavior
`paper`	Simulated against live prices. No wallet or chain activity. Default.
`dry-run`	Builds and logs each step but does not submit transactions.
`confirm-each`	Submits live orders, prompting for confirmation before each step.
`auto-execute`	Submits live orders without prompting. Use with guardrails.

Margin mode

For TWAP and grid runs, --margin-mode selects how collateral is held:

Mode	Behavior
`cross`	Uses the trader’s main cross-margin account. Default.
`isolated`	Opens an isolated subaccount. Pair with `--isolated-collateral` for the opening deposit.

Guardrails

All three runners accept the same safety flags. Any breach pauses the run and records the reason in the ledger:

Flag	Description
`--max-total-notional-usdc <amount>`	Cap on total live notional placed by the run.
`--max-step-notional-usdc <amount>`	Cap on a single tick’s live notional.
`--max-price-drift-bps <bps>`	Pause if mark price drifts this far from the run-start price.
`--max-exposure-ratio <ratio>`	Pause if position notional / equity exceeds this ratio.
`--reconcile-attempts <n>`	Number of history reconciliation attempts per live step.
`--reconcile-delay-ms <ms>`	Milliseconds between reconciliation attempts.

TWAP

TWAP slices a target size across N equally spaced intervals so each fill nudges the average price toward the time-weighted mean.

`vulcan strategy twap start`

Flag	Description
`--symbol <s>`	Market symbol (e.g. `SOL`).
`--side <buy\|sell>`	Trade direction.
`--notional-usdc <amount>`	Total TWAP notional in USDC. Mutually exclusive with `--tokens`.
`--tokens <amount>`	Total TWAP size in base-asset tokens. Mutually exclusive with `--notional-usdc`.
`--slices <n>`	Number of slices.
`--interval-seconds <s>`	Seconds between slices. Default `60`.
`--mode <m>`	Execution mode. Default `paper`.
`--margin-mode <m>`	Margin mode. Default `cross`.
`--isolated-collateral <amount>`	USDC to transfer when opening the first isolated live order.
`--run-label <label>`	Optional human-readable label stored with the run.
`--detached`	Start the runner in the background and return immediately with the run ID.

Plus the guardrail flags above.

`vulcan strategy twap resume`

vulcan strategy twap resume <run-id> [--from-step <n>]

Resume a paused or incomplete TWAP run, optionally starting at a specific step.

Example

vulcan strategy twap start \
  --symbol SOL \
  --side buy \
  --notional-usdc 5000 \
  --slices 10 \
  --interval-seconds 300 \
  --mode auto-execute \
  --max-step-notional-usdc 600 \
  --max-price-drift-bps 75 \
  --detached

Grid

Grid trading lays buy levels below the mark and sell levels above the mark, then maintains the ladder as fills happen. The runner can either be bounded by --ticks or run indefinitely with --run-until-stopped.

`vulcan strategy grid start`

Flag	Description
`--symbol <s>`	Market symbol (e.g. `SOL`).
`--lower-price <p>`	Lower grid boundary price. Required unless `--center-on-mark` is set.
`--upper-price <p>`	Upper grid boundary price. Required unless `--center-on-mark` is set.
`--center-on-mark`	Center the grid on the current mark price at launch. Requires `--width-pct`.
`--width-pct <pct>`	Half-width of the grid as a percentage of mark (e.g. `1.0` = ±1%). Requires `--center-on-mark`.
`--levels-per-side <n>`	Number of buy levels below mark and sell levels above mark.
`--tokens-per-level <t>`	Order size per level in base-asset tokens. Mutually exclusive with `--size-lots-per-level`.
`--size-lots-per-level <n>`	Order size per level in base lots. Mutually exclusive with `--tokens-per-level`.
`--bid-level PRICE:SIZE_LOTS[:TP][:SL]`	Fully custom bid level. Repeatable, replaces generated bids.
`--ask-level PRICE:SIZE_LOTS[:TP][:SL]`	Fully custom ask level. Repeatable, replaces generated asks.
`--take-profit-spacing <p>`	Distance from entry to take-profit for generated levels.
`--stop-loss-spacing <p>`	Distance from entry to stop-loss for generated levels.
`--interval-seconds <s>`	Seconds between maintenance ticks. Default `60`.
`--ticks <n>`	Maximum ticks to run, including the initial placement tick. Default `60`.
`--run-until-stopped`	Keep running maintenance ticks until paused or stopped.
`--stale-after-seconds <s>`	Seconds without a tick before status reports this run as stale.
`--mode <m>`	Execution mode. Default `paper`.
`--margin-mode <m>`	Margin mode. Default `cross`.
`--isolated-collateral <amount>`	USDC to transfer when opening the first isolated live order.
`--run-label <label>`	Optional human-readable label.
`--slide`	Allow live multi-limit orders to slide to the top of book if crossing.
`--detached`	Start in the background and return immediately with the run ID.

Plus the guardrail flags above.

`vulcan strategy grid resume`

vulcan strategy grid resume <run-id> [--from-step <n>]

Resume a paused or incomplete grid run.

Example

vulcan strategy grid start \
  --symbol SOL \
  --center-on-mark \
  --width-pct 2.5 \
  --levels-per-side 5 \
  --tokens-per-level 0.5 \
  --run-until-stopped \
  --mode auto-execute \
  --max-total-notional-usdc 10000 \
  --detached

TA

The TA runner evaluates rules of the form (condition, action) over indicator values on a chosen timeframe. Conditions reference indicators like rsi, macd, ema, etc.; actions place or close positions when triggered. Configs are JSON, supplied inline or via a file. TA margin settings are part of that JSON config (margin_mode and isolated_collateral), not separate CLI flags.

`vulcan strategy ta start`

Flag	Description
`--config-file <path>`	Path to a JSON config file. Mutually exclusive with `--config-json`.
`--config-json <json>`	Inline JSON config. Mutually exclusive with `--config-file`.
`--mode <m>`	Execution mode. Default `paper`.
`--max-ticks <n>`	Maximum ticks to run. Ignored when `--run-until-stopped` is set. Default `60`.
`--run-until-stopped`	Keep running until paused or stopped.
`--run-label <label>`	Optional human-readable label.
`--detached`	Start in the background and return immediately with the run ID.

Plus the guardrail flags above.

`vulcan strategy ta resume`

vulcan strategy ta resume <run-id>

Resume a paused or incomplete TA strategy run.

Example

vulcan strategy ta start \
  --config-file ./ema-cross-sol.json \
  --mode paper \
  --run-until-stopped \
  --detached

For ad-hoc indicator queries outside the strategy framework, use the top-level vulcan ta commands.

Lifecycle commands

These apply to any strategy run, regardless of type. Most take the <run-id> returned by start (runs and preflight do not). List recent run IDs with vulcan strategy runs.

Command	Description
`runs [--limit <n>]`	List persisted strategy runs. Default limit 20.
`status <run-id> [--since-tick <n>] [--include-ledger]`	Show latest status. `--since-tick` returns only ticks newer than the given index.
`monitor <run-id> [--include-ledger]`	Compact non-blocking monitor state.
`wait-next-tick <run-id> [--after-tick <n>] [--timeout-seconds <s>] [--include-ledger]`	Block until a new tick or terminal status is observed. Default timeout 90 seconds.
`report <run-id>`	Show the final (or latest) report.
`reconcile-grid <run-id>`	Inspect live grid orders against the persisted ledger without mutating state.
`pause <run-id> [--reason <r>]`	Request the runner to pause at the next safe point.
`stop <run-id> [--reason <r>]`	Request the runner to stop permanently at the next safe point.
`finalize <run-id> [--reason <r>] [--cancel-orders] [--close-position] [--wait] [--timeout-seconds <s>]`	Stop the runner and optionally cancel open orders and close the position on the strategy symbol.
`preflight`	Pre-launch readiness check: wallet identity, password availability, trader registration, collateral. Lists every blocker with a remedy command.
`resume <run-id> [--from-step <n>]`	Resume any paused or incomplete strategy run.

Typical operating loop

RUN_ID=$(vulcan strategy grid start ... --detached -o json | jq -r '.data.run_id')

vulcan strategy monitor "$RUN_ID" -o json
vulcan strategy wait-next-tick "$RUN_ID" --timeout-seconds 120

vulcan strategy pause "$RUN_ID" --reason "checking risk"
vulcan strategy resume "$RUN_ID"

vulcan strategy finalize "$RUN_ID" --cancel-orders --close-position --wait

Overview

Reference

Execution modes

Margin mode

Guardrails

TWAP

`vulcan strategy twap start`

`vulcan strategy twap resume`

Example

Grid

`vulcan strategy grid start`

`vulcan strategy grid resume`

Example

TA

`vulcan strategy ta start`

`vulcan strategy ta resume`

Example

Lifecycle commands

Typical operating loop

​Execution modes

​Margin mode

​Guardrails

​TWAP

​vulcan strategy twap start

​vulcan strategy twap resume

​Example

​Grid

​vulcan strategy grid start

​vulcan strategy grid resume

​Example

​TA

​vulcan strategy ta start

​vulcan strategy ta resume

​Example

​Lifecycle commands

​Typical operating loop

Execution modes

Margin mode

Guardrails

TWAP

`vulcan strategy twap start`

`vulcan strategy twap resume`

Example

Grid

`vulcan strategy grid start`

`vulcan strategy grid resume`

Example

TA

`vulcan strategy ta start`

`vulcan strategy ta resume`

Example

Lifecycle commands

Typical operating loop