Reports with charts

Numbers are useful when they explain what is happening. These charts keep the work readable without turning it into a pitch deck.

Notebook snapshot: 2026-05-18.

Follow the review workflow

Start with a question, learn the term, open the risk review overview, then come here for the numbers. The notes make more sense in motion.

What is moving

The current catalog tracks 8 systems, 21 source entries, 26 number notes, and 35 visible limits.

The learning layer adds 20 concepts and 15 guide steps, so the numbers do not float without language around them.

What this is not

This is not employer data, client data, live trading performance, or a compliance certification.

It is a report: visible records, numbers that can be explained, and limits that keep the story honest.

Systems 5 lanes
Source entries 14 entries
Number notes 16 notes
Limits named 21 boundaries
Learning concepts 20 concepts
Guide steps 15 steps

What sits behind each section

Each section shows how many links, numbers, limits, and tags sit behind it. This is a quick notebook map, not a ranking.

Wallet Risk Engine + Assessment Notes 5 links / 5 numbers / 5 limits / 5 tags
Agent Workflow System 2 links / 3 numbers / 4 limits / 4 tags
Agent Chess 3 links / 3 numbers / 4 limits / 4 tags
Monitoring Heartbeat 2 links / 4 numbers / 4 limits / 4 tags
Agent Notes Kit 2 links / 2 numbers / 4 limits / 3 tags
Kite Testnet Proof 2 links / 4 numbers / 5 limits / 4 tags
Agent Operations Lessons 2 links / 2 numbers / 4 limits / 4 tags
Fraud Alert Triage Workflow 3 links / 3 numbers / 5 limits / 5 tags

Numbers worth keeping close

These values come from existing summaries. Bars use a display scale per metric family, so percentages, function counts, tests, and prediction counts stay readable without pretending they measure the same thing.

These numbers come from test runs, not live trading. Each metric has a method note that explains what was measured and what the limit is.

Wallet-risk benchmark detection 94.9 percent. From: Wallet-risk engine summary.
Wallet-risk benchmark false-positive rate 20.2 percent. From: Wallet-risk engine summary.
Direct detection functions 23 functions. From: Wallet-risk rules manifest summary.
Mapped detection abilities 28 abilities. From: Wallet-risk rules manifest summary.
Agent framework local tests 70 tests. From: Agent Framework summary.
Heartbeat predictions 2514 predictions. From: Heartbeat snapshot.

Where the records feel strong or thin

This is a practical map. Wallet-risk notes, agent coordination, and observability are already readable. Deployment is the next visible question.

Wallet risk 5/5
Agent governance 5/5
Observability 4/5
Public learning 4/5
Deployment 2/5

How to read these numbers

Counts are derived from visible catalogs, not every note in the workspace.

Benchmark values are only shown where they already exist in a source summary.

Limits are treated as a feature, not a weakness, because they keep the story honest.

The lightest area is packaging. A small reproducible demonstration would make the next pass easier to follow.

Get the Bionic source trail

One short note when an AI, finance, crypto, or risk signal becomes worth documenting. Source-backed notes only. No trading calls, no hype.