[feature] add optional slippage simulation for paper market orders

[theghostmac]

Context

Market orders in paper trading filled at the exact best ask/bid with zero slippage, causing paper PnL to consistently overstate what a strategy would achieve live -- especially for larger order sizes, illiquid pairs, and volatile conditions.

This PR adds an optional --slippage flag to paper init and paper reset, following existing --fee-rate pattern.

Brief Changelog

• Added slippage_rate: f64 to PaperState with #[serde(default)] for backward compat with existing state.json files
• Added DEFAULT_SLIPPAGE_RATE = 0.0 constant
• Extended with_fee_rate and reset_with to accept slippage_rate
• Market buys now fill at ask * (1 + slippage_rate), sells at bid * (1 - slippage_rate)
• Limit orders unaffected
• Added --slippage CLI flag to paper init and paper reset
• Added slippage rate to both human-readable and JSON output
• Added tests covering buy slippage, sell slippage, zero slippage backward compat, and reset preservation/update

Usage:

Update: flags --slippage-rate or --slippage works.

Init with slippage rate:

./target/debug/kraken paper init --balance 10000 --slippage-rate 0.001
┌──────────────────┬───────────────────────────┐
│ Field            ┆ Value                     │
╞══════════════════╪═══════════════════════════╡
│ Mode             ┆ [PAPER] Simulated Trading │
│ Action           ┆ Account initialized       │
│ Starting Balance ┆ 10000.00 USD              │
│ Fee Rate         ┆ 0.26%                     │
│ Slippage Rate    ┆ 0.10%                     │
└──────────────────┴───────────────────────────┘

Reset slippage rate:

./target/debug/kraken paper reset --slippage-rate 0.002
┌──────────────────┬───────────────────────────┐
│ Field            ┆ Value                     │
╞══════════════════╪═══════════════════════════╡
│ Mode             ┆ [PAPER] Simulated Trading │
│ Action           ┆ Account reset             │
│ Starting Balance ┆ 10000.00 USD              │
│ Fee Rate         ┆ 0.26%                     │
│ Slippage Rate    ┆ 0.20%                     │
└──────────────────┴───────────────────────────┘

Closes #15

[agent-smith-k]

Thanks for the contribution — the approach is solid and mirrors the existing --fee-rate pattern well. Core slippage math is correct and tests cover the happy paths. A few items to address before merge:

Must fix

1. Typo in JSON output key (paper init).
   "slipage_rate" (one p) on line 238 of src/commands/paper.rs.
   paper reset correctly uses "slippage_rate". Any -o json consumer
   will see inconsistent keys depending on the command. One-character fix.

2. Add a serde backward-compatibility test.
   The current zero_slippage_is_backward_compatible test uses PaperState::new(),
   which always sets slippage_rate via the constructor. It does not prove that an
   older state.json (without the field) deserializes correctly. Please add a test
   similar to deserialize_state_missing_next_order_id that parses JSON without
   slippage_rate and asserts it defaults to 0.0.

Should fix

3. Missing semicolon in validation block.
   In execute_reset, the slippage validation return Err(...) omits the trailing
   semicolon, while the fee_rate validation block directly above it includes one.
   Both compile, but the inconsistency is noticeable. Add ; for consistency.

4. Serde default style mismatch.
   fee_rate uses #[serde(default = "default_fee_rate")] with a named function.
   slippage_rate uses bare #[serde(default)], relying on f64::default() == 0.0.
   This works today, but if the default ever changes to nonzero, the serde default
   would silently diverge from DEFAULT_SLIPPAGE_RATE. Consider adding a
   default_slippage_rate function for consistency.

5. No tests for validation edge cases.
   Both execute_init and execute_reset validate slippage bounds
   (finite, 0.0..=1.0), but there are no tests exercising those paths.
   Worth adding at least: negative rate, rate > 1.0, NaN, and Infinity.

Clarification

6. Flag name: --slippage vs --slippage-rate.
   The issue text proposes --slippage, but this PR exposes --slippage-rate.
   --slippage-rate is arguably more precise (matches --fee-rate). Either way
   is fine, but please align the issue description or add --slippage as a
   clap alias so both work.

Non-blocking design notes (for follow-up)

7. with_fee_rate is becoming a positional-params bag.
   It now takes four positional args, two of which are f64 and easy to swap.
   The name no longer describes the function. Not blocking, but a config struct
   or builder would prevent misuse as more options get added.

8. paper status does not show slippage rate.
   init and reset display it, but status does not. A user who set slippage
   weeks ago has no way to check the value without reading state.json directly.
   Good candidate for a follow-up.

Items 1 and 2 should be fixed before merge. The rest are improvements that would be appreciated but won't block on.

[theghostmac]

Thank you @agent-smith-k. Addressed both blocking and non-blocking.

[agent-smith-k]

All findings from the initial review have been addressed. The JSON key typo is fixed, serde defaults are consistent, validation is extracted into a reusable helper with edge-case tests, backward-compat deserialization is tested, --slippage alias works, paper status now surfaces both rates, and PaperConfig eliminates the positional-params concern.

296/296 tests pass. Runtime verification confirms correct JSON output for both paper init and paper status.

One low-severity hardening note for a follow-up: persisted state.json values are trusted on reload without re-validation. Not blocking since the trust boundary is local.

Approving.

[agent-smith-k]

Superseded by #24, which contains your changes as a signed squash commit with Co-authored-by credit. Merged to main. Thank you for the contribution!