Curriculum·J306 Smart Contract Risk for Non-Developers·about 31 min
A risk rubric
By the end of this lesson you can
- →Score a protocol across the six dimensions this course produces
- →Apply the minimum rule rather than an average, and justify why
- →Identify the gate conditions that disqualify regardless of other scores
- →Convert a rubric result into a position size rather than a verdict
Junior · enrolled learners
This lesson opens with The average that was always high.
- What happened
- This is a pattern drawn from the verified incidents in this course and this level rather than a single event, and it is labeled as such. Multichain was a cross-chain bridge holding enough user assets that more than $125M left it in July 2023, in a loss Chainalysis read as compromised keys rather than faulty code; by the team's own account, the servers running its MPC nodes were under its chief executive's personal cloud account. Venus's May 2021 loss ran through one listing decision: it let users borrow against its own governance token, XVS, and was left with more than $95M of bad debt. Compound had a governance process that worked as designed and, by its founder's account, could not change anything in less than seven days. Helio accepted Ankr's aBNBc as collateral; when an attacker minted trillions of aBNBc and its price crashed, Helio's oracle had not updated, and $16M of its HAY stablecoin was borrowed against the collapsed token. In every case, in this course's judgment, a scorecard averaging across dimensions would have returned a high number, and in every case a single dimension determined the outcome.
- The decision point
- Averaging is the wrong operation for a system whose failure requires only one path to be open. Per J301-04, security is the cost of the cheapest authorization path rather than the typical one, and a rubric that averages six dimensions will rate a protocol scoring nine on five of them and two on the sixth at 7.83, which is a number that describes nothing about what will happen. The correct summary statistic for a stack of dependencies is the minimum.
What you will be able to answer
- →What are the six rubric dimensions?
- →Why the minimum rather than the average?
- →What are the gate conditions?
- →What does the rubric produce?
Orientation and Year One are open: anyone can read them without an account. From Year Two onward the lessons are for enrolled learners, because progress through the later years only means anything if it is tracked against a record.
It is free. We do not sell the list and there is nothing to buy at the end of it.
Sources and review
- https://www.theblock.co/news/ecosystems/2023-07-14-multichain-ceases-operations-as-ceos-sister-is-detained-in-china-while-holding-220-million-of-funds-239589
- https://www.chainalysis.com/blog/multichain-exploit-july-2023/
- https://cointelegraph.com/news/multichain-stops-operations-over-lack-of-funds
- https://www.theblock.co/news/defi/2021-05-19-bsc-venus-protocol-liquidations-xvs-token-possible-price-manipulation-105301
- https://www.theblock.co/news/defi/2021-09-29-compound-bug-comp-risk-misreward-119086
- https://www.theblock.co/news/regulation/2022-12-02-attacker-pockets-20-million-in-exploits-on-ankr-and-helio-191668
- https://www.ankr.com/blog/after-action-report-our-findings-from-abnbc-token-exploit/
Confidence high·Volatility low·Reviewed 2026-10-22·Owner unassigned
Contested
The pattern in the autopsy is this lesson's synthesis, not a finding of any single source. Each case is sourced on its own, in the lesson of this course or level that covers it; the list above carries The Block (14 July 2023) and Chainalysis for Multichain, The Block for Venus (19 May 2021, which reports that Venus let users borrow against XVS and was left with more than $95 million of bad debt) and for Compound, and The Block (2 December 2022, citing BlockSec) and Ankr's own after-action report for Ankr and Helio. The cloud account detail is the Multichain team's own account, published while its chief executive was in custody, and has not been independently confirmed. The claim that an averaging scorecard would have rated each one highly is a judgment this course makes from those cases, and it is labeled as such.
The size bands in the worked example are illustrative and depend entirely on a reader's own policy per S207-06. The transferable part is the mapping from a minimum score to a bound, not the specific percentages.
J310-07 applies this rubric to tokenized real world assets, where the dimensions extend to a legal wrapper, an issuer's jurisdiction and a redemption process running on business days. J308-02 owns governance as an attack surface. Both extend this rubric rather than replacing it.