Skip to main content
Glama

Fmi

by io.github.pipeworx-io

Polymarket Edges

polymarket_edges
Read-onlyIdempotent

Scan top Polymarket markets and return opportunities where Pipeworx data disagrees with market price. Built for "what should I bet on today" — agents discover opportunities without paging hundreds of markets. FIVE MODEL FAMILIES grouped into three response segments under by_segment: (1) MODEL_DRIVEN — crypto_price (lognormal barrier from 90d FRED log-returns) and news_momentum (GDELT 7d/21d article-volume ratio, soft signal w/ halved Kelly). (2) STRUCTURAL_ARBITRAGE — partition_overround on mutually-exclusive events; per-leg favorite-longshot bias correction with per-sport α (tennis 1.02, soccer 1.10, MMA 1.15, default 1.0); placeholder-slug filter drops will-person-X / will-team-Y / will-manager-Z / will-someone-else- backstops; partitions with >20% placeholder fraction skipped entirely. (3) CONCENTRATED_LONGSHOT — basket trade when one leg ≥75% AND ≥2 longshots ≤8% AND portfolio return ≥25:1; rare-by-design (gates relaxed Run 8 from prior 85%/5%/50:1). EVERY OPPORTUNITY carries edge_pp_net (after slippage), kelly_fraction + kelly_fraction_half (capped at 0.25), market.liquidity, market.spread_pp, market.volume, plus a 24h-move warning ("Market moved X.Xpp in 24h") when the recent move alone exceeds the edge — your edge may already be in the price. TRADEABLE-EDGE KNOBS: min_liquidity / max_spread_pp drop opportunities where edge isn't realizable; min_partition_leg_kelly filters partitions by best per-leg Kelly. RESPONSE TOP-LEVEL: by_segment{model_driven,structural_arbitrage,concentrated_longshot}, fed_candidates/fed_note (Fed bets surface here, excluded from ranking — 1m-T vs EFFR signal is unreliable at meeting-month horizons without paid OIS/SOFR-futures data), and _diagnostics{concentrated_longshot:{...funnel counters},category_counts,filter_skips} so callers can see WHY a segment is empty (top-N stale, all candidates failed gates, knob dropped them). Cached 1h at the KV level keyed on all knobs.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
limitNoTop N edges to return after ranking. Default 10, max 25.
windowNoPolymarket volume window to filter markets. Default 1wk.
min_kellyNoMinimum half-Kelly fraction (as decimal, e.g. 0.005 = 0.5% of bankroll) to include single-leg opportunities. Default 0 (no filter). Skips opportunities that are too small to bet sensibly even if the edge is large.
min_edge_ppNoMinimum |edge| in percentage points to include (default 0.5). Edge is evaluated NET of slippage.
slippage_ppNoAssumed execution slippage in percentage points per leg (default 0.3). Subtracted from raw |edge| before ranking and Kelly sizing. Polymarket has zero trading fees as of 2024 but bid/ask + thin depth typically eats 20-50bp per trade. Bump for very thin partitions; drop to 0 if you have a smarter fill model.
max_spread_ppNoTradeable-edge filter. Maximum bid/ask spread in percentage points on the representative market. Default null (no filter). Set to 2 to require tight books — anything wider eats most plausible edges.
min_liquidityNoTradeable-edge filter. Minimum $ liquidity on the representative market (or for partition_overround, on at least one top_leg). Default 0 (no filter). Set to 5000 to drop thin-book opportunities where executing the edge would walk the book past breakeven.
category_filterNoComma-separated list to restrict the output: "model_driven" (crypto_price + news_momentum), "structural_arbitrage" (partition_overround), "concentrated_longshot". Combine like "model_driven,structural_arbitrage". Default: all.
min_partition_leg_kellyNoMinimum BEST per-leg half-Kelly fraction across a partition_overround opportunity's top_legs (or longshot_basket legs). Default 0 (no filter). Partition arbs always return kelly_fraction_half=0 at the parent level by design (basket trades don't compose to single-leg Kelly), so min_kelly never filters them — this knob applies to the per-leg Kelly inside top_legs instead. Use to suppress thin partitions whose individual leg edges aren't worth the per-leg slippage cost.

TDQS

A4.7/5.0
Behavior5/5

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

Even though annotations already declare readOnlyHint=true and destructiveHint=false, the description adds substantial behavioral context: the cache behavior ('Cached 1h at the KV level keyed on all knobs'), response diagnostics (_diagnostics so callers can see WHY a segment is empty), per-model details (placeholder-slug filter, per-sport bias correction α), and valuation cautions (the 24h-move warning, Fed note about unreliable signal). It also explains the 'rare-by-design' nature of concentrated longshots. No contradiction with annotations.

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 description is long but well-structured, using capital-label sections (MODEL_DRIVEN, STRUCTURAL_ARBITRAGE, CONCENTRATED_LONGSHOT, TRADEABLE-EDGE KNOBS, RESPONSE TOP-LEVEL) that aid navigation. It is front-loaded with the core purpose. While every sentence carries substantive information, the length is high; however, given the tool's complexity (9 optional parameters, 4 model variants, nested response structure), the detail is justified rather than padded.

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 carries the full burden of explaining the return structure, and it does so thoroughly: top-level keys (by_segment, fed_candidates/fed_note, _diagnostics), per-opportunity fields (edge_pp_net, kelly_fraction, market.liquidity, market.spread_pp), and even the funnel counters in diagnostics for empty segments. It covers model logic, filters, knobs, and caching, making it fully adequate for an agent to understand what it will receive and why.

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

Parameters5/5

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

Schema description coverage is 100%, so baseline is 3, but the description significantly enriches parameter meaning. It labels min_liquidity and max_spread_pp as 'tradeable-edge filters', explains that min_partition_leg_kelly applies to per-leg Kelly inside top_legs because 'partition arbs always return kelly_fraction_half=0 at the parent level by design', and gives practical tuning advice (e.g., 'Bump for very thin partitions; drop to 0 if you have a smarter fill model'). These details go far beyond generic schema descriptions, clarifying parameter intent and interactions.

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?

The description opens with a clear, specific verb-resource pair: 'Scan top Polymarket markets and return opportunities where Pipeworx data disagrees with market price.' It names the exact resource (Polymarket markets), the action (scan and return opportunities), and the core purpose (finding mispricings against Pipeworx data). It also explicitly frames the use case ('what should I bet on today'), distinguishing it from siblings like polymarket_arbitrage or polymarket_edge_tracker.

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 provides clear usage context: 'Built for "what should I bet on today" — agents discover opportunities without paging hundreds of markets.' It explains the intended scenario and the value proposition. However, it does not explicitly name alternative tools for exclusion (e.g., 'for tracking edge over time use polymarket_edge_tracker') or mention when not to use this tool, so it falls short of a 5.

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

All Fmi toolsInspect
Try in Browser
RayrunBuild, deploy, and operate MCP servers without becoming an infrastructure team.ray.runAd

Glama MCP Gateway

Add one secure layer between your agents and this server.

TDQS

A3.9/5.0
Disambiguation4/5

Most tools are well-differentiated, with detailed descriptions clarifying their distinct purposes. However, there is some overlap between the multiple 'ask' tools (ask_pipeworx, ask_pipeworx_grounded, deep_research, suggest_questions) and between weather tools (forecast, latest_observations, recent_observations, warnings), which could cause confusion. Overall, ambiguity is low.

Naming Consistency5/5

All 34 tool names follow a consistent lowercase_with_underscores (snake_case) pattern. Names are descriptive and predictable, such as 'ask_pipeworx', 'entity_profile', 'polymarket_arbitrage', etc. No mixing of conventions like camelCase or inconsistent verb styles.

Tool Count3/5

The server has 34 tools, which is on the higher side given its broad scope covering weather, company research, prediction markets, and general data queries. While not excessive, it could be split into more focused servers for clarity. The count feels a bit heavy but still manageable.

Completeness4/5

The tool set covers a wide range of data domains (weather, company financials, prediction markets, news, memory, subscriptions) with reasonable completeness. Minor gaps exist, such as limited weather coverage (Finland only) and no direct support for non-company entities or unofficial data sources, but core workflows are well-supported.

Resources