Skip to main content
Glama

FirmTape - SPX dealer positioning

Test an options strategy on the archive

options_strategy_test

Test an SPX 0DTE options strategy — a condor, strangle, straddle or any leg set — against every session in the FirmTape archive on real quotes and real settlement. One call prices it 63 ways: seven widths of your geometry across nine entry times, so the answer is WHERE TO PUT THE STRIKES rather than a verdict on one guess. Short legs are sold at the bid and long legs bought at the ask, $1.25 a leg — measured on 13.1M prints, the median SPX 0DTE print pays the full quoted half-spread, so these fills are the median outcome and not a conservative one. Returns the marginal by width and by entry, a recommended direction ONLY when the marginal is ordered and a withheld holdout agrees, the day-clustered mean and its error, prob_loss_20, and all 63 cells with yours marked. If those 63 come back flat the run carries on by itself for the same credits -- where the structure sits relative to spot, past the edge of the grid, then eight standard 0DTE shapes -- up to 154 configurations, so the answer is 'not this, and here is what the archive does pay for' rather than a shrug. It never names a best cell: the best of 63 is best by chance. Spends 10 credits, refunded on failure. Requires a Lab token (Authorization header).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
legsYesThe strategy, as legs. `dist` is the leg's distance from spot as a fraction: -0.005 is half a percent below, +0.0083 is 0.83% above. Strikes are not given because they travel — the geometry is what the archive is asked. A short 0.5% strangle is two legs at -0.005 and +0.005, both qty -1.
entryNoYour entry time. All nine are reported whatever you pick; this only marks which cell is yours.

TDQS

A4.4/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Discloses material behavior beyond annotations: it costs 10 credits, refunds on failure, requires a Lab token, uses median bid/ask fills at $1.25 per leg, reports a recommended direction only under strict conditions, never names a best cell, and may auto-expand to 154 configurations. No contradiction with the annotations; readOnlyHint=false is consistent with a credit-spending test operation.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The purpose is front-loaded and the detail is mostly earned given the tool's complexity, covering output shape, fill assumptions, costs, and failure behavior. Some phrasing is verbose and slightly winding, but no critical information is buried.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema, the description properly enumerates returned items: marginal by width and entry, a conditional recommended direction, the day-clustered mean and error, prob_loss_20, and the 63 marked cells. It also covers pricing, cost, refunds, authentication, and the expanded grid behavior, making it thorough enough to invoke correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema already provides 100% descriptive coverage for legs, qty, dist, right, and entry, including clear examples. The description adds context about the 63-way grid and that entry only marks which cell is yours, but it does not substantially expand parameter-level meaning beyond what the schema already gives.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: 'Test an SPX 0DTE options strategy' against the FirmTape archive on real quotes and real settlement. This clearly differentiates it from the sibling data-retrieval tools like get_gex_profile, get_levels, and get_session, which do not backtest strategies.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description makes it clear this is for testing options strategies against archived real-market data and describes the kind of answer it produces, so an agent can infer when to choose it over the alternatives. However, it does not explicitly name sibling tools or state when not to use it.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

A4.6/5.0
Disambiguation5/5

Every tool targets a distinct resource and granularity: documents (search/fetch), session enumeration (list_sessions), day-level measurements (get_session/get_levels), intraday shape (get_gex_profile), multi-day trends (get_level_history), and strategy backtesting (options_strategy_test). The descriptions go further than usual, using explicit 'Not for' cross-references so even get_levels vs get_session — a deliberate strict subset — reads unambiguously.

Naming Consistency4/5

The measurement tools follow a clean get_<nouon> convention (get_session, get_levels, get_level_history, get_gex_profile) and list_sessions fits the verb_nouon pattern. fetch and search are bare single-word verbs that break the pattern slightly — fetch_document / search_documents would be more uniform — and options_strategy_test is a compressed compound, but the overall snake_case style and semantic grouping are predictable.

Tool Count5/5

Eight tools is well within the 3-15 sweet spot for a specialized data server. Each tool carries a clearly differentiated job: two document access, one enumeration, four measurement surfaces (single-day full, single-day levels, intraday ladder, multi-day history), and one analysis operation. Nothing feels like filler and nothing essential is missing.

Completeness4/5

The domain is well covered: session discovery (list_sessions), document retrieval (search/fetch), full daily readings (get_session), subset views (get_levels), intraday gamma shape (get_gex_profile), and cross-day movement (get_level_history) — plus a strategy tester. Minor gaps exist, such as no way to compare gamma-ladder shape across multiple days without repeated get_gex_profile calls, but no workflow dead-ends for the stated purpose.

Resources