Tools reference

Trinity MCP tools

The 12 Trinity MCP tools are read-only, atemporal and tier-aware. No tool mutates user state, sends an order, or triggers an action outside of a read request on the Trinity knowledge base.

list_metrics

Freeread-only

Paginated, tier-gated index of the Trinity library. Each entry carries the minimal identifiers to drill in: metric_id, rubric, name, required tier, category, catalog number, exclusive flag. Only metrics your tier can access are returned (no upgrade nag).

Parameters

Name	Type	Required	Description
rubric	`string`	no	Filter by rubric: onchain, macro, cycle, price, narrative, exchange, technical.
search	`string`	no	Full-text keyword (max 64 chars). Composes with rubric.
limit	`integer`	no	Results per page (1 to 200, default 50). Paginate via offset.
offset	`integer`	no	Number of entries to skip before returning results (default 0).
include_descriptions	`boolean`	no	true to include each entry's FR/EN description (default false, keeps the response under the 25K-token ceiling).
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: metrics[] (metric_id, rubric, name, tier_required, is_exclusive, catalog_number, category), pagination (total_matching, returned, limit, offset, next_offset), filter_applied, compliance.

get_chart_value

Free (per chart)read-only

Fetches a certified Trinity chart value: extrema-preserving downsampled history (every cycle top and bottom is guaranteed present), latest value, embedded EDCD education and TAE audit status. Access checked dynamically against your tier.

Parameters

Name	Type	Required	Description
metric_id	`string`	yes	Chart slug (e.g. mvrv-z-score), from list_metrics.
timeframe	`string`	no	Window: 7d, 30d, 90d, 1y, 2y, 5y, all. Default 1y.
asset	`string`	no	BTC, ETH or SOL. Default BTC.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: metric_id, name, timeframe, asset, data_points_count, data[] (extrema-preserving downsampled), sampling (method + counts), latest, education, audit_status, metadata (tier_required, certified_version, freshness_hours).

Note: For an exclusive composite (TCTI, Trinity Code…), only the score is returned - the formula is never disclosed. Downsampling mathematically preserves each bucket's min and max (cycle extremes are precisely what matters).

get_chart_education

Free (per chart)read-only

Fetches a Trinity chart's full EDCD educational context, atemporal and bilingual: what it is, how to read, key zones, what to observe, historical context, common mistakes.

Parameters

Name	Type	Required	Description
metric_id	`string`	yes	Trinity chart slug.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: metric_id, name, education (whatIsIt, howToRead, keyZones, whatToObserve, historicalContext, commonMisinterpretations), is_exclusive, certified_version, compliance.

get_audit_status

Freeread-only

Fetches a metric's Trinity Audit Engine (TAE) certification status. The TAE compares the Trinity value daily against a public reference industry source and certifies institutional fidelity.

Parameters

Name	Type	Required	Description
metric_id	`string`	yes	Chart slug to audit.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: metric_id, audit_status (pass | warn | fail | no_audit), delta_pct vs external source (FRED, ECB, CoinMetrics, Blockstream), compliance.

Note: Exclusive Trinity composites return no_audit (no external industry comparable).

compare_metrics_across_rubrics

Founding / Proread-only

Compares 2 to 5 metrics from different Trinity rubrics (on-chain, macro, cycle, narrative, price, exchange) over a date range. EXCLUSIVE Trinity capability. E.g. 'does BTC narrative attention lead MVRV peaks?'

Parameters

Name	Type	Required	Description
metrics	`array<string>`	yes	2 to 5 metric_ids from any rubric.
date_range	`string`	yes	Format YYYY-MM-DD:YYYY-MM-DD.
comparison_mode	`string`	no	correlation, alignment or lead-lag. Default correlation.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: cross-rubric reading (correlation / alignment / lead-lag) + atemporal interpretation, without revealing any formula, + compliance.

compare_cycles

Performanceread-only

Overlays 2 to 4 Bitcoin cycles aligned on halving (or another normalization) for a given metric. Trinity holds 4 complete cycles in a unified database (2009-2013, 2013-2017, 2017-2021, 2021-2024+).

Parameters

Name	Type	Required	Description
cycles	`array<integer>`	yes	2 to 4 cycle anchor years.
metric	`string`	yes	Metric slug to overlay.
normalize	`string`	no	days_since_halving, days_since_cycle_bottom or absolute. Default days_since_halving.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: overlaid series per cycle, current cycle phase position, atemporal interpretation, compliance.

get_trinity_composite

Founding / Proread-only

Fetches the score (0-100) and phase of a proprietary Trinity composite (TCTI, Trinity Code, TLBI, TSS, TSMI, TRI, TAI). The formula, weightings and inputs are NEVER disclosed.

Parameters

Name	Type	Required	Description
composite_name	`string`	yes	TCTI, Trinity Code, TLBI, TSS, TSMI, TRI or TAI. "TCI" is still accepted for backward compatibility: it returns no value and a consolidation notice pointing to TSMI.
date_range	`string`	no	Optional, format YYYY-MM-DD:YYYY-MM-DD.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: 0-100 score + phase (0-25 accumulation / early bull, 25-50 mid bull, 50-75 late bull, 75-100 distribution / cycle top) + atemporal interpretation + compliance.

Note: Recipe-leak guard: only the bounded score and phase are returned. Founding / Pro tier required.

get_trinity_indicator

Per indicatorread-only

Fetches the latest score (0-100) and history of a proprietary Trinity v3 indicator (the indicators behind the /indicators dashboard: TCTI, TRI, TSMI, TSS, TMHS…). Distinct from catalog charts.

Parameters

Name	Type	Required	Description
key	`string`	yes	Lowercase key (e.g. tcti, tri, tsmi, tss, tglc). List via /api/v1/indicators.
limit	`integer`	no	History points (1 to 5000, default 30).
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: key, name, category, tier_min, latest (score, regime, direction, confidence, timestamp, metadata), history[] (oldest to newest), compliance.

Note: Recipe-leak guard: the composition (inputs, weights) is never returned - only the public surface. Tier = indicator_meta.tier_min.

get_glossary_term

Freeread-only

Fetches the Trinity definition of a term (TCTI, EDCD, TAE, MVRV, NUPL, SOPR, NVT, RHODL, Liveliness, Dormancy…). Bilingual, atemporal, aligned with chart educational content.

Parameters

Name	Type	Required	Description
term	`string`	yes	Term to define (2 to 32 chars).
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: term, definition (in the requested language), language, compliance.

get_hyperliquid_ecosystem

Freeread-only

Free aggregate snapshot of one of the seven Hyperliquid sections: market-pulse, edge-explorer, institutional-lab, hip3, hip4, etf, hype. Returns ecosystem-level numbers (open interest, volume, counts, ETF, HYPE composite score).

Parameters

Name	Type	Required	Description
section	`string`	yes	market-pulse, edge-explorer, institutional-lab, hip3, hip4, etf or hype.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: section, snapshot (ecosystem numbers, read-only HYPE composite score), compliance.

get_hyperliquid_history

Founding / Proread-only

Time-series drill-down of a Hyperliquid metric (sections market-pulse, hip3, hip4, etf, hype). Pro+ tier required. The identifier targets a row (ETF ticker, HIP-4 outcome_id).

Parameters

Name	Type	Required	Description
section	`string`	yes	market-pulse, hip3, hip4, etf or hype.
metric	`string`	yes	Metric column to plot (list via the trinity://hyperliquid/sections resource).
period	`string`	no	7d, 30d, 90d, 1y or all. Default 30d.
identifier	`string`	no	ETF ticker (required for section=etf) or HIP-4 outcome_id (optional).
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: time-series of the requested metric + compliance.

Note: Recipe-leak guard: the HYPE composite sub-factors are never queryable - metric='composite_score' returns the score only.

get_hyperliquid_exclusive

Performanceread-only

Trinity-exclusive Hyperliquid lecture (Performance tier): HLP counter-trade reading, clearinghouse health pointer, HYPE composite snapshot. Score, phase and atemporal interpretation only.

Parameters

Name	Type	Required	Description
view	`string`	yes	hlp-counter-trade, clearinghouse-health or hype-composite-snapshot.
language	`string`	no	fr or en. Default: fr.

Returned payload

JSON object: score + phase + atemporal interpretation + compliance.

Note: Recipe-leak guard: each lecture's composition (factors, weights, smoothing, inputs) is NEVER disclosed. Performance tier required.