Tested statistics of 179 chart and candlestick patterns on six markets, against random bars.
Scored across 4 tools
Each tool has a clearly distinct purpose: searching/listing patterns, retrieving full stats for one pattern, comparing multiple patterns, and explaining methodology. There is little risk of confusing get_pattern_stats with search_patterns or compare_patterns, since their descriptions explicitly distinguish single-pattern detail, filtered listing, and side-by-side comparison.
All four tools follow a consistent snake_case verb_noun or verb_object pattern: search_patterns, get_pattern_stats, compare_patterns, get_methodology. The convention is predictable and readable throughout.
Four tools is well-scoped for a read-only pattern-statistics server. Search, detail, comparison, and methodology each earn their place without redundancy.
The surface covers discovery, detailed retrieval, comparison, and methodology documentation, which are the core workflows for this domain. A minor gap exists because patterns recorded on both breakout sides are excluded from search results and can only be retrieved via get_pattern_stats if the name is already known, and there is no dedicated tool for enumerating markets/timeframes beyond methodology documentation.
Put 2 to 10 patterns side by side on the same market, timeframe and breakout side.
| Name | Required | Description | Default |
|---|---|---|---|
| side | No | Breakout side. Patterns recorded on both sides have up and down rows and no all row; every other pattern has only all. Default: all. | |
| market | No | Market. all pools the markets that have this timeframe. Default: all. | |
| patterns | Yes | Pattern ids, slugs or names. | |
| timeframe | No | Bar timeframe. Default: 1d. Stocks, ETFs, indices and commodities have 1d and 1w only. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already establish this as a safe read (readOnlyHint=true, openWorldHint=false), so the bar is lower. The description does add one useful behavioral fact: all compared patterns resolve against one shared market/timeframe/side filter. It says nothing about what the comparison output contains or ordering, leaving a real gap for a multi-entity operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single sentence, front-loaded with the verb and the 2-10 bound, with zero filler or redundant restatement of the title. Every clause carries information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema, the description is the only place to learn what a side-by-side comparison returns (metrics, table shape, ordering), and it stays silent. The input side is adequately covered, but the result side is not, which matters for a tool whose whole purpose is producing a comparison.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, with enums and defaults documented inline for side, market, and timeframe, so the schema already carries the parameter meaning. The description repeats the shared market/timeframe/side framing without adding syntax or constraints beyond the schema; baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Names a specific verb ("Put ... side by side") and resource (patterns), and the constraint of a shared market/timeframe/side makes the operation concrete. It does not explicitly contrast with siblings like get_pattern_stats or search_patterns, which is the only thing keeping it from a 5.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The 2-to-10 pattern range implies a multi-pattern comparison scenario versus single-pattern lookups, so usage is inferable. However, there is no explicit when-to-use, no mention of when a single-pattern tool such as get_pattern_stats is the better choice, and no prerequisites stated.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
How the statistics are computed: principles, the meaning of every field and verdict, and data coverage (assets and date range per market and timeframe).
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations only declare readOnlyHint=true and openWorldHint=false; the description goes further by enumerating the actual content returned (principles, field meanings, verdicts, per-market/per-timeframe coverage), which is the behavior an agent needs to know. It does not state the return format (text vs. structured), a minor gap for a zero-parameter reference tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single front-loaded sentence with a colon-delimited inventory of contents; nothing is redundant and the scope is stated before any detail.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema and no parameters, the description carries the burden of telling the agent what comes back, and it does so by listing the four content areas. Only the response format/size is left unspecified, which is a small residual gap for a documentation tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool takes no parameters, so the baseline is 4. The description correctly implies the full methodology and coverage matrix is returned unconditionally with no filtering options.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description names the resource precisely: documentation of how the statistics are computed, including principles, field/verdict meanings, and data coverage. That is clearly distinct from the analytical siblings (compare_patterns, get_pattern_stats, search_patterns), though it never explicitly states the contrast.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Usage is only implied: an agent infers it should call this to interpret fields and verdicts returned by the stats tools, or to check coverage per market/timeframe. There is no explicit 'call this before/after X' statement and no naming of alternatives to avoid.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Full tested results of one pattern: detection rule, trade rules and every market, timeframe and breakout side measured, with confidence intervals, q-values, verdicts, timing quartiles and the page URL to cite.
| Name | Required | Description | Default |
|---|---|---|---|
| side | No | Keep only this breakout side. Default: every side. | |
| market | No | Keep only this market. Default: every market. | |
| pattern | Yes | Pattern id (e.g. engulfing_bullish), page URL or slug, or name (e.g. "bullish engulfing"). | |
| timeframe | No | Keep only this timeframe. Default: every timeframe. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and openWorldHint=false, so the safety profile is covered. The description adds real value by enumerating the returned statistics (confidence intervals, q-values, verdicts, timing quartiles, citation URL), but says nothing about failure modes for an unresolved pattern id or response size, so it is useful context rather than rich behavioral disclosure.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
A single sentence with the purpose front-loaded and the return contents following. It is dense but every clause names concrete data the agent will get; the colon-separated enumeration is slightly list-heavy but not padded.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With no output schema, the description must describe what comes back, and it does so concretely down to q-values, timing quartiles, and a citable page URL. The main remaining gap is error behavior when the pattern identifier does not resolve.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100% and enums are supplied for side, market, and timeframe, so the schema carries the parameter burden. The phrase 'every market, timeframe and breakout side measured' reinforces that these are filter dimensions over a full result set, but adds no format or default semantics beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific resource and scope: 'Full tested results of one pattern', enumerated as detection rule, trade rules, and per-market/timeframe/side measurements. The 'one pattern' framing implicitly sets it apart from compare_patterns and search_patterns, though no sibling is named explicitly.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
No when-to-use statement, no exclusions, and no reference to the obvious alternatives (search_patterns to find a pattern, compare_patterns for multi-pattern comparison). The single-pattern scope is only implied by the title and description text.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List tested patterns with their headline result on one market and timeframe. Filter by name, family, direction or verdict, for example: which chart patterns beat chance on forex 4h. Patterns recorded on both breakout sides have no all row and are not listed: read them with get_pattern_stats.
| Name | Required | Description | Default |
|---|---|---|---|
| sort | No | occurrences (default), excess (excess return, best first), trade (expectancy minus its random baseline, best first) or name. | |
| limit | No | Maximum number of patterns returned. Default: 50. | |
| query | No | Words of the pattern name, e.g. "engulfing", "head and shoulders". | |
| family | No | Pattern family (harmonic patterns are in chart). | |
| market | No | Market. all pools the markets that have this timeframe. Default: all. | |
| measure | No | move: excess 10-bar return over the rest of the market (verdict). trade: 2:1 trade versus the same trade on random dates (verdictTrade). Default: move. | |
| verdict | No | Keep only this verdict (on the measure chosen with measure). | |
| direction | No | Expected direction. | |
| timeframe | No | Bar timeframe. Default: 1d. Stocks, ETFs, indices and commodities have 1d and 1w only. | |
| min_occurrences | No | Keep only results with at least this many occurrences. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnlyHint=true and openWorldHint=false, so safety is covered; the description adds a real behavioral caveat beyond that — patterns recorded on both breakout sides lack an "all" row and are excluded from results. That is exactly the kind of silent omission an agent would otherwise misread as absent data.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Three sentences, front-loaded with what is returned, then a filter summary plus example, then the exclusion rule. Every sentence carries information the agent needs; nothing is padding.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a 10-parameter, zero-required search tool with no output schema, the description conveys the row shape at a high level ("headline result") and the key coverage caveat. It stops short of describing what columns or verdict fields come back, which would help given there is no output schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the enum meanings and defaults are already documented in structured fields. The description reinforces that results are pinned to a single market and timeframe but adds no syntax, unit, or interaction detail beyond the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Names a specific verb and resource ("List tested patterns") plus the scope of the returned row ("headline result on one market and timeframe"). It also distinguishes itself from get_pattern_stats by explaining which patterns are omitted here and where to read them.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Gives a concrete use case ("which chart patterns beat chance on forex 4h") that maps directly onto the family/market/timeframe/verdict parameters, and states an explicit carve-out routing the agent to get_pattern_stats. It does not distinguish itself from compare_patterns, which is a remaining gap.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
compare_patternsget_methodologyget_pattern_statssearch_patternsBreakout patterns on stocks & crypto, each with its historical win rate, entry, stop and target.
Verified Japanese equity studies: 50 anomalies, 26 indicators, 12 candlesticks, events, ETF decay.
Praxis: published trading strategies re-run and audited — verdicts, claimed vs measured stats.
Crypto chart past cases: same coin, same exchange shape search with what followed each match.
Add one secure layer between your agents and this server.
