dhawalorg
Agent · MachineWhat can I ask or reuse?

Ask a sourced question or open the same evidence as data.

MCPAgent-native infrastructureImmutable claim archive · as of 2026-07-30

The machine floor.

The desk publishes the exact state, analog, signal, and thesis payload behind each daily claim. Every payload has a permanent URL and SHA-256 hash. Aggregate tool traffic remains public without recording who asked or what they typed.

Choose your surface

This is the audit trail.

Most readers do not need hashes to use the desk. This page is for developers, researchers, and agents who want to verify the exact archived evidence behind a claim.

Live protocol proof

One desk. Two readers.

This browser sends a real MCP tools/call to the public endpoint. The returned regime is composed from the same sourced macro and breadth routes used by the human desk.

POST /mcptools/callget_regime/api/macro + /api/breadth
01 · requestJSON-RPC 2.0
{
  "jsonrpc": "2.0",
  "id": "machine-floor-regime",
  "method": "tools/call",
  "params": {
    "name": "get_regime",
    "arguments": {}
  }
}
02 · responseQueued after first paint
Macro regime 73.1/100 · risk-on
5 of 5 legs reporting · as of 2026-07-29
Source: FRED and Yahoo Finance
Method: Equal-weight mean of the finite volatility, breadth, credit, 2s10s, and dollar leg scores.
HTTP pendingTiming pending
Raw JSON-RPC response
Waiting for the live response. The archived evidence remains visible above.
Number73.1 / 100
Observation2026-07-29 · n=5/5 legs
SourceFRED and Yahoo Finance
MethodFive-leg composite
Machine Floor v2 · as of 2026-07-30

Every claim leaves evidence.

The mutable latest files are conveniences. The evidence record is content-addressed and immutable. Recompute SHA-256 over the exact JSON payload and compare it with the printed hash.

MCP tools
23 · n=23
Claim snapshots
32 · n=32
Theses archived
32 · n=32
Falsifiers evaluated
31 · n=32
get_state_vector2026-07-30

State Vector

Number
25 dimensions · n=25
Sample
4,924 dated sessions · n=4924
Formula
Each current reading carries its component formula, prior-window z-score, percentile, and sample size.
SHA-256 payload
9cc151862de6e847a16e4dd3aac3c514c952728b76de49d15fff94964e8038ac
find_analogs2026-07-30

Analog Engine

Number
50 matched states · n=50
Sample
4,840 eligible sessions · n=4840
Formula
Cosine distance over shared finite dimensions after each date is standardized against only its prior 252 sessions.
SHA-256 payload
3d6807db0bab6ec0ed58b7ad9536a8d48a95c276f5940b04bf3172a708e29977
get_signal_performance2026-07-30

Signal Performance

Number
7 definitions · n=7
Sample
1 published from n=7 audited definitions
Formula
Walk-forward folds, 21-session independence, publication floor n=50.
SHA-256 payload
2d661c6a2527b46a8bef9cae8e7b579a53f23e2a95832b44200f8fb910d26327
Public accountability

The desk keeps its own score.

Each thesis keeps the original anomaly, conditional history, falsifier, exact evidence payloads, and the first eligible print used to decide whether the falsifier fired.

2026-07-30

creditHygLqdRatio

PENDING · n=0

HYG/LQD printed 0.745999, 2.60 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 4 disagree, and 3 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.92% median, 62.0% positive hit rate, and -2.33% worst decile; ES had a +0.69% median, 62.0% positive hit rate, and -2.40% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725891, the z=0.00 level from the prior n=252 observations, on its next daily print.

No eligible observation after 2026-07-30. Evaluation sample n=0.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741319, 2.26 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 56.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742426, 2.45 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.56% worst decile; ES had a +0.87% median, 58.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 56.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741125, 2.23 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 58.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741321, 2.26 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 58.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 56.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a -0.10% median, 48.0% positive hit rate, and -2.75% worst decile; ES had a -0.29% median, 48.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a -0.10% median, 48.0% positive hit rate, and -2.75% worst decile; ES had a -0.29% median, 48.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741125, 2.23 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 58.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a -0.10% median, 48.0% positive hit rate, and -2.75% worst decile; ES had a -0.29% median, 48.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a -0.10% median, 48.0% positive hit rate, and -2.75% worst decile; ES had a -0.29% median, 48.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.740517, 2.15 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a -0.10% median, 48.0% positive hit rate, and -2.75% worst decile; ES had a -0.29% median, 48.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-17

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741297, 2.26 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.64% median, 56.0% positive hit rate, and -2.75% worst decile; ES had a +0.81% median, 58.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725445, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.74363 on 2026-07-20. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742494, 2.46 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

1 components agree with the risk-on direction, 3 disagree, and 5 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.68% median, 56.0% positive hit rate, and -2.56% worst decile; ES had a +0.64% median, 56.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742886, 2.52 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.99% median, 60.0% positive hit rate, and -2.56% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742518, 2.46 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.56% worst decile; ES had a +0.87% median, 58.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742531, 2.46 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.56% worst decile; ES had a +0.87% median, 58.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742425, 2.45 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.56% worst decile; ES had a +0.87% median, 58.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742426, 2.45 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.56% worst decile; ES had a +0.87% median, 58.0% positive hit rate, and -2.83% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725403, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-16

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.740517 on 2026-07-17. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.743455, 2.72 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.75% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725244, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.741866, 2.40 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 58.0% positive hit rate, and -2.78% worst decile; ES had a +1.11% median, 60.0% positive hit rate, and -2.91% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725360, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.743455, 2.72 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.75% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725244, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.743455, 2.72 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.75% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725244, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.743455, 2.72 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.75% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725244, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-15

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.743455, 2.72 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

2 components agree with the risk-on direction, 3 disagree, and 4 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +0.89% median, 58.0% positive hit rate, and -2.75% worst decile; ES had a +1.02% median, 60.0% positive hit rate, and -2.85% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725244, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.742326 on 2026-07-16. Evaluation sample n=1.

2026-07-14

creditHygLqdRatio

HELD · n=1

HYG/LQD printed 0.742853, 2.58 standard deviations above its prior 252-observation mean, the largest absolute state-vector z-score.

0 components agree with the risk-on direction, 3 disagree, and 6 have no vote. Component sample n=9.

Across n=50 analogs, SPX +21-session returns had a +1.01% median, 60.0% positive hit rate, and -2.32% worst decile; ES had a +1.11% median, 62.0% positive hit rate, and -2.10% worst decile.

Falsifier

The read is invalid if HYG/LQD records at or below 0.725310, the z=0.00 level from the prior n=252 observations, on its next daily print.

Observed 0.741866 on 2026-07-15. Evaluation sample n=1.

This edge · UTC window2026-07-30
Tool calls0
Successful·
Connecting
Live aggregate

Watch the desk answer machines.

One rail per canonical agent tool, plus any specialist tool called in this window. Width is the edge-local call count. The best-effort counter resets at UTC midnight and may reset earlier if the edge cache is evicted.

ask_desk0
0 ok0 failedNo calls in this window
explain_signal0
0 ok0 failedNo calls in this window
find_analogs0
0 ok0 failedNo calls in this window
get_13f0
0 ok0 failedNo calls in this window
get_brief0
0 ok0 failedNo calls in this window
get_calendar0
0 ok0 failedNo calls in this window
get_congress_trades0
0 ok0 failedNo calls in this window
get_dataset0
0 ok0 failedNo calls in this window
get_quote0
0 ok0 failedNo calls in this window
get_regime0
0 ok0 failedNo calls in this window
get_signal_performance0
0 ok0 failedNo calls in this window
get_state_vector0
0 ok0 failedNo calls in this window
get_symbol0
0 ok0 failedNo calls in this window
run_screen0
0 ok0 failedNo calls in this window
Collected

Tool name, success or failure, aggregate timestamps.

Structurally excluded

Arguments, prompts, symbols, IP addresses, user agents, sessions.

Call the same desk

The interface is public.

Connect to https://dhawal.org/mcp, read the agent skill, or inspect the OpenAPI contract. Every result keeps its source, formula, sample size, immutable evidence URL, and payload hash attached.