defi-invariant-check
Skill220 starsChanged 20 days agoMITView on GitHub
Use when the audit target is a DeFi protocol — AMM, lending market, yield vault, stablecoin, or perps — and you must author the economic properties generic campaigns miss, such as token-conservation, constant-product k, share accounting, collateralization, and oracle-deviation invariants, then falsify them.
In this repository179 other files it gives its agents.
- AGENTS.mdAGENTS.md
- chrono/CLAUDE.mdCLAUDE.md
- CLAUDE.mdCLAUDE.md
- accessible-media-authoringSkill
- agentic-safety-auditSkill
- agent-prompt-engineeringSkill
- audio-event-map-authoringSkill
- auto-scaffoldSkill
- chain-constructSkill
- chain-construct-smart-contractSkill
- claim-verificationSkill
- code-reachability-auditSkill
- code-review-loopSkill
- color-theorySkill
- conversation-designSkill
- copy-refinementSkill
- cosmos-sdk-audit-checklistSkill
- cross-chain-bridge-auditSkill
- cross-file-relationship-synthesisSkill
- data-flow-traceSkill
- dedup-prior-art-checkSkill
- dependency-cycle-auditSkill
- dependency-health-triageSkill
- detection-as-codeSkill
- diff-aware-semgrep-scanSkill
- differential-reviewSkill
- dimensional-analysis-checkSkill
- dual-level-retrievalSkill
- durable-nonce-exploitationSkill
- erc1271-revert-data-checkSkill
- error-based-sstiSkill
- evm-audit-flowSkill
- figma-implement-designSkill
- findings-filterSkill
- forensic-timeline-authoringSkill
- game-design-fundamentalsSkill
- game-mechanics-balancingSkill
- head-tailSkill
- incident-response-runbookSkill
- interactive-audio-designSkill
- interface-ambiguity-checkSkill
- keyword-clusteringSkill
- knowledge-base-integrationSkill
- known-advisory-backport-checkSkill
- layered-analysis-loopSkill
- level-design-patternsSkill
- locale-adaptationSkill
- multi-stance-audit-fanoutSkill
- narrative-structureSkill
- osint-platform-auditSkill
- parser-differential-route-confusionSkill
- persuasive-technical-writingSkill
- platform-complianceSkill
- player-engagement-psychologySkill
- pre-audit-threat-modelSkill
- program-rubric-lookupSkill
- read-only-reentrancy-checkSkill
- requirements-elicitationSkill
- review-severity-ladderSkill
- rule6-rights-gateSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 3 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 26
Only authorized roles can set oracle addresses, fee parameters, or pause state.
- Distinctive: line 79
No re-entrant call can increase the caller's balance beyond their pre-call entitlement.
- Very specific: line 85
Author Echidna properties: Write properties as Solidity functions with prefix echidna_ that return bool.
- Very specific: line 91
Use testMode: assertion in echidna.yaml for complex multi-step violations.
- Very specific: line 126
Do NOT author invariants that pass by construction — e.g., an echidna_ property that trivially returns true because the function under test …
- Very specific: line 127
Do NOT treat a single Echidna passed result as proof of correctness — Echidna is a fuzzer, not a prover.
- Distinctive: line 128
Do NOT skip flash-loan simulation for lending/AMM protocols — flash loans invalidate many invariants that hold under normal call conditions.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
<!-- inspired by trailofbits/skills (concept); repointed to native CLIs for Chrono -->defi-invariant-check
DeFi-specific invariant authoring and testing discipline. Use this as a sub-skill during
evm-audit-flow when the target is a DeFi protocol (AMM, lending, yield, bridge, stablecoin,
perps). Generic Echidna/Medusa campaigns miss DeFi-specific invariants — this skill provides
the property checklist. All tools are the native host CLIs (echidna, medusa, forge,
halmos) run in the target project directory; there is no wrapper layer.
Invariant categories
Token accounting invariants (always required):
totalSupply == sum(balanceOf(all_holders))— no spurious minting or burning.reserve0 * reserve1 >= k(for constant-product AMMs) — k-invariant never decreases except via fee accrual.totalBorrow <= totalDepositfor all lending pools — protocol cannot lend out more than deposited.sum(userShares) == totalSharesfor yield vaults — share accounting is consistent.
Access-control invariants:
- Only authorized roles can set oracle addresses, fee parameters, or pause state.
onlyOwner/onlyGovernancefunctions cannot be called by arbitrary external accounts.- Upgradeable proxy implementation slots cannot be modified except through the upgrade pathway.
Economic/oracle invariants:
- Flash loan callbacks cannot leave the protocol in a net-loss state after the loan repayment.
- Liquidation proceeds must cover the borrowed amount plus protocol fee — no bad-debt accrual from liquidations.
2026 incident-derived oracle-input robustness invariants (highest-priority economic class):
- Spot versus TWAP: a price used for collateral, borrowing, liquidation, minting, or
redemption must reject or clamp a same-block spot price when
abs(spot - TWAP) / TWAP > maxDeviation; no value-affecting path may silently fall back to the divergent spot. This targets the oracle/valuation-manipulation class represented by Moonwell (2026-08-27), Lazy Summer (2026-07-06), Makina (2026-01-20), and Edel (2026-07-01) inV114-51. - Low-liquidity collateral: within the protocol's configured manipulation budget, an
atomic trade against a thin market cannot increase an account's borrowable value or
liquidation proceeds by more than the attacker's net cost plus the configured safety
margin. Reject collateral whose observable depth cannot support the valuation notional.
This is the Moonwell low-liquidity-collateral oracle shape from
V114-51. - NAV manipulation bound: absent external profit or loss, a deposit/reprice/redeem round
trip across any supported asset must not return more value than entered after fees and
rounding tolerance, and one manipulable component price must not move total NAV beyond
its configured exposure and deviation caps. This targets the Lazy Summer, Makina, and
Edel valuation/NAV incident class from
V114-51.
2026 incident-derived governance/timelock invariants (second priority):
- Delay cannot be bypassed or shortened: for every queued privileged operation,
executedAt - queuedAt >= max(delayAtQueue, protocolMinimumDelay). Changing the delay, executor, proposer, or role graph after queueing cannot accelerate that operation, and a delay change itself becomes effective only after the pre-change delay. This targets Term Finance's 2026-08-24 timelock-zeroing shape fromV114-51. - A role or executor change cannot authorize its own proposal or execution; every privilege
expansion must be queued and executed by the pre-change authority after the full
pre-change delay. Snapshot the audited role graph and assert that no post-deployment role
drift broadens value-moving authority outside that path. This targets the TOP
(2026-06-26) and BonkDAO (2026-07-26) governance-bypass shapes from
V114-51.
Rounding/precision invariants (third priority, still required):
- Exercise zero-supply initialization, one-unit deposits/withdrawals, values immediately below and above every division boundary, and repeated dust operations. No non-zero asset input may mint zero shares unless the transaction reverts or the loss is explicitly bounded; no positive supply or assets may exist without corresponding claimable shares.
- A deposit/redeem or mint/withdraw round trip, with no external profit or loss, must return the starting economic value within the documented fee and rounding bound; splitting one operation into many dust-sized operations cannot improve that bound.
- Fuzz solver domains and every fixed-point math-library boundary used by share conversion,
including zero denominators, scale changes, and minimum/maximum supported values. These
properties retain the dust/zero-supply class seen twice at Thetanuts in June 2026, while
placing it behind the two larger 2026 code-level classes identified by
V114-51.
Reentrancy invariants:
- Protocol state (balances, reserves, shares) after any external call must equal the state written before the call (CEI pattern verification).
- No re-entrant call can increase the caller's balance beyond their pre-call entitlement.
Order of ops
- Identify the protocol class. AMM, lending, vault, bridge, stablecoin, or perps. Each class has a canonical invariant set (use the category checklist above).
- Author Echidna properties. Write properties as Solidity functions with prefix
echidna_that returnbool. Properties must be pure invariants (no state mutations inside the property). Place intest/invariants/EchidnaTest.sol. - Run Echidna — minimum 30 minutes for DeFi protocols:
Useechidna . --contract EchidnaTest --config echidna.yaml --test-limit 500000testMode: assertioninechidna.yamlfor complex multi-step violations. - Run Medusa as a second pass if Echidna does not find violations — Medusa's
coverage-guided corpus often finds violations Echidna's random mutation misses:
medusa fuzz --config medusa.json - Flash-loan attack simulation. Manually construct a flash-loan attack call sequence in a
Forge test that instantiates a
FlashLoanAttackercontract, then runforge test --match-contract FlashLoanAttacker -vvvv. Verify the k-invariant or balance-conservation invariant holds after the attack sequence. - Price-oracle and NAV manipulation check. For protocols with on-chain price oracles:
write a Forge test that varies market depth, creates large single-block spot/TWAP
divergence, and reprices each NAV component. Check whether the oracle-derived price would
enable profitable liquidations, under-collateralized borrows, or a profitable
deposit/reprice/redeem round trip (
forge test --match-test testOracleManip -vvvv). - Governance/timelock check. Queue each value-moving privileged operation, then mutate the delay and role graph and attempt early or self-authorized execution. Use the target framework's time-warp primitive to assert that execution remains impossible until the greater of the queued delay and protocol minimum has elapsed.
- Rounding/precision check. Fuzz zero-supply initialization, dust values around every division boundary, split-versus-batched operations, and deposit/redeem round trips. Keep tolerances in the protocol's documented units; a tolerance must not erase a zero-share mint or ghost-supply state.
When to pivot
- Echidna cannot construct valid state transitions: the protocol requires complex setup (governance votes, time-locks, external oracle seeding). Author a setup harness in a
CryticSetupcontract that initializes the protocol to a valid post-deployment state. - Invariant is too weak: Echidna falsifies it immediately with trivial inputs. Tighten the property — e.g., add a precondition (
require(totalSupply > 0)) to avoid vacuous falsification. - Protocol uses proxy pattern: Echidna tests the proxy ABI; ensure the proxy's
fallbackroutes correctly to the implementation in the test environment.
Anti-patterns
- Do NOT author invariants that pass by construction — e.g., an
echidna_property that trivially returnstruebecause the function under test is never called. - Do NOT treat a single Echidna
passedresult as proof of correctness — Echidna is a fuzzer, not a prover. Usehalmosfor bounded formal guarantees on critical properties (halmos --function check_ --loop 4). - Do NOT skip flash-loan simulation for lending/AMM protocols — flash loans invalidate many invariants that hold under normal call conditions.
Example
Writing and running an Echidna token conservation invariant:
// test/invariants/EchidnaTest.sol
contract EchidnaTest {
Token token;
function echidna_total_supply_equals_sum() public view returns (bool) {
// totalSupply must equal sum of all holder balances
return token.totalSupply() == token.balanceOf(address(this))
+ token.balanceOf(address(0xdead));
}
}echidna . --contract EchidnaTest --config echidna.yaml --test-limit 500000Recording (chrono-vault)
The task packet's injected memory contract owns the exact call shape, sequence, and fields - see
wirework-reflect. Do not copy a record(...) example or add fields (including source_task) from
memory; the server binds them, and a baked example violates the run's authenticated schema. Memory is
best-effort telemetry and never gates the work. What is worth recording here is the task-specific
outcome: protocol class, invariants tested, violations, flash-loan/oracle-manip tested?.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

