VegaSec / Solidity scan report

VegaTokenVulnerable.sol

VSS-20260924-61NL6Q44 · generated Sep 24, 2026, 4:42 PM

AVOID-style triage14 findings1 critical3 highPasted source1 warningsemgrep 1.163.0

Verdict is derived from severity counts via the same deterministic formula every VegaSec report uses. VegaTokenVulnerable.sol was scanned from pasted source. This is static-analysis triage, not a certification or formal audit attestation.

Signature

signed

This report is signed. Anyone can verify that these exact bytes came from us, with no account, no wallet and no API key. The signature proves integrity and authorship. It does not prove the audited contract is safe: read the coverage status and the limitations above.

sha256
4757b05e2d549932...93c0075c
Signed by
Ed25519 key vsig-2026-09
Canonical version
vss-canonical-1
Bytes served
stored canonical text, byte for byte

PDF artifact

Queued artifact pipeline · not blocking on the scan

Not requested

No PDF artifact has been requested for this report yet. The HTML report and the Markdown export already carry the full disclaimer and limitations inline. PDF synthesis is a queued artifact, not part of the scan path.

Status endpoint

GET /report/solidity/VSS-20260924-61NL6Q44/export.pdf

Solidity evidence report

14 audit findings

critical: 1high: 3medium: 6low: 1informational: 3

Severity distribution

Executive summary

  • 14 findings surfaced (1 critical, 3 high, 6 medium, 1 low, 3 informational).
  • CRITICAL · selfdestruct call — exploitability unknown.
  • HIGH · VegaTokenVulnerable.unsafeReviewerLottery(address[]) (VegaTokenVulnerable.sol#28-34) uses a weak PRNG: "picked = reviewers[block.timestamp % reviewers.length] ( (unsafeReviewerLottery) — exploitability medium.
  • HIGH · Reentrancy in VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42): (withdrawReviewCredit) — exploitability medium.
  • Automated static analysis only. A human operator review has not been attached to this report.

Warnings (1)

  • · Unmapped detector slither:immutable-states — classified as 'other'
semgrep 1.163.0
  • criticalCentralizationConfidence: unknownExploitability: unknownDetector: semgrep

    selfdestruct call

    Affected location
    VegaTokenVulnerable.sol:47
    Detector / source
    semgrep · semgrep:vegasec-selfdestruct

    Evidence

    selfdestruct call. Even with access control, selfdestruct introduces upgrade risk and historical state loss. Audit access control carefully or remove.

    Exploitability notes

    Analyzer did not report a confidence level; treat as unverified until reviewed.

    Remediation

    Document the trust boundary, set up time-locked multisig ownership, and surface admin-only functions in the deployment runbook.

    Limitations

    • Analyzer did not report confidence; fail closed and treat the finding as unverified.
    • Static analysis did not prove exploitability; do not present this as an exploit claim.
    • Analyzer output did not include a source excerpt for this finding.
  • highOracle manipulationConfidence: mediumExploitability: mediumDetector: slither

    VegaTokenVulnerable.unsafeReviewerLottery(address[]) (VegaTokenVulnerable.sol#28-34) uses a weak PRNG: "picked = reviewers[block.timestamp % reviewers.length] (

    Affected location
    VegaTokenVulnerable.sol:28-34 · unsafeReviewerLottery
    Detector / source
    slither · slither:weak-prng

    Evidence

    VegaTokenVulnerable.unsafeReviewerLottery(address[]) (VegaTokenVulnerable.sol#28-34) uses a weak PRNG: "picked = reviewers[block.timestamp % reviewers.length] (VegaTokenVulnerable.sol#31)"

    Exploitability notes

    Medium-confidence detection; manual review required before relying on the finding either way.

    Remediation

    Use multi-source price feeds (e.g. Chainlink + TWAPs) and validate freshness/deviation before consuming oracle data on-chain.

    Limitations

    • Medium analyzer confidence; manual review should confirm the code path and preconditions.
    • Exploitability depends on deployment context, permissions, balances, and reachable preconditions.
    • Analyzer output did not include a source excerpt for this finding.
  • highReentrancyConfidence: mediumExploitability: mediumDetector: slither

    Reentrancy in VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42):

    Affected location
    VegaTokenVulnerable.sol:36-42 · withdrawReviewCredit
    Detector / source
    slither · slither:reentrancy-eth

    Evidence

    Reentrancy in VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42): External calls: - (ok,None) = address(msg.sender).call{value: amount}() (VegaTokenVulnerable.sol#39) State variables written after the call(s): - balances[msg.sender] -= amount (VegaTokenVulnerable.sol#41) VegaTokenVulnerable.balances (VegaTokenVulnerable.sol#10) can be used in cross function reentrancies: - VegaTokenVulnerable.balances (VegaTokenVulnerable.sol#10) - VegaTokenVulnerable.constructor() (VegaTokenVulnerable.sol#15-18) - VegaTokenVulnerable.mintForReview(address,uint256) (VegaTokenVulnerable.sol#22-26) - VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42)

    Exploitability notes

    Medium-confidence detection; manual review required before relying on the finding either way.

    Remediation

    Apply the checks-effects-interactions pattern. State writes must happen before external calls, and consider OpenZeppelin's ReentrancyGuard for high-risk entry points.

    Limitations

    • Medium analyzer confidence; manual review should confirm the code path and preconditions.
    • Exploitability depends on deployment context, permissions, balances, and reachable preconditions.
    • Analyzer output did not include a source excerpt for this finding.
  • highAccess controlConfidence: highExploitability: highDetector: slither

    VegaTokenVulnerable.retireFixture() (VegaTokenVulnerable.sol#44-48) allows anyone to destruct the contract

    Affected location
    VegaTokenVulnerable.sol:44-48 · retireFixture
    Detector / source
    slither · slither:suicidal

    Evidence

    VegaTokenVulnerable.retireFixture() (VegaTokenVulnerable.sol#44-48) allows anyone to destruct the contract

    Exploitability notes

    High-confidence detection of a severe pattern; treat as exploitable until disproven by manual review.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • High analyzer confidence, still requires source-level review before promotion.
    • Exploitability is inferred from a severe high-confidence static signal, not from a live exploit proof.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumAccess controlConfidence: mediumExploitability: mediumDetector: slither

    VegaTokenVulnerable.mintForReview(address,uint256) (VegaTokenVulnerable.sol#22-26) uses tx.origin for authorization: require(bool,string)(tx.origin == owner,ori

    Affected location
    VegaTokenVulnerable.sol:22-26 · mintForReview
    Detector / source
    slither · slither:tx-origin

    Evidence

    VegaTokenVulnerable.mintForReview(address,uint256) (VegaTokenVulnerable.sol#22-26) uses tx.origin for authorization: require(bool,string)(tx.origin == owner,origin owner only) (VegaTokenVulnerable.sol#24)

    Exploitability notes

    Medium-confidence detection; manual review required before relying on the finding either way.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • Medium analyzer confidence; manual review should confirm the code path and preconditions.
    • Exploitability depends on deployment context, permissions, balances, and reachable preconditions.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumAccess controlConfidence: unknownExploitability: unknownDetector: semgrep

    Use of tx.origin for authorization

    Affected location
    VegaTokenVulnerable.sol:24
    Detector / source
    semgrep · semgrep:vegasec-tx-origin-auth

    Evidence

    Use of tx.origin for authorization. tx.origin can be tricked via phishing contracts. Prefer msg.sender.

    Exploitability notes

    Analyzer did not report a confidence level; treat as unverified until reviewed.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • Analyzer did not report confidence; fail closed and treat the finding as unverified.
    • Static analysis did not prove exploitability; do not present this as an exploit claim.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumTimestamp dependenceConfidence: unknownExploitability: unknownDetector: semgrep

    block.timestamp used in a value computation

    Affected location
    VegaTokenVulnerable.sol:31
    Detector / source
    semgrep · semgrep:vegasec-block-timestamp-rng

    Evidence

    block.timestamp used in a value computation. Miners can adjust block.timestamp within ~15s; do not use it for randomness or time-critical comparisons.

    Exploitability notes

    Analyzer did not report a confidence level; treat as unverified until reviewed.

    Remediation

    Treat block.timestamp as a coarse value with miner influence; prefer block-number windows for short-term invariants and add tolerance bands.

    Limitations

    • Analyzer did not report confidence; fail closed and treat the finding as unverified.
    • Static analysis did not prove exploitability; do not present this as an exploit claim.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumUnchecked callConfidence: unknownExploitability: unknownDetector: semgrep

    Low-level .call to an externally-controlled target

    Affected location
    VegaTokenVulnerable.sol:39
    Detector / source
    semgrep · semgrep:vegasec-low-level-call-untrusted

    Evidence

    Low-level .call to an externally-controlled target. Combined with reentrancy or unchecked return values, this is a common fund-draining pattern.

    Exploitability notes

    Analyzer did not report a confidence level; treat as unverified until reviewed.

    Remediation

    Check the boolean return of low-level call/delegatecall/send and revert on failure with a descriptive reason string.

    Limitations

    • Analyzer did not report confidence; fail closed and treat the finding as unverified.
    • Static analysis did not prove exploitability; do not present this as an exploit claim.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumAccess controlConfidence: mediumExploitability: mediumDetector: slither

    VegaTokenVulnerable.retireFixture() (VegaTokenVulnerable.sol#44-48) uses tx.origin for authorization: require(bool,string)(tx.origin == owner,origin owner only)

    Affected location
    VegaTokenVulnerable.sol:44-48 · retireFixture
    Detector / source
    slither · slither:tx-origin

    Evidence

    VegaTokenVulnerable.retireFixture() (VegaTokenVulnerable.sol#44-48) uses tx.origin for authorization: require(bool,string)(tx.origin == owner,origin owner only) (VegaTokenVulnerable.sol#46)

    Exploitability notes

    Medium-confidence detection; manual review required before relying on the finding either way.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • Medium analyzer confidence; manual review should confirm the code path and preconditions.
    • Exploitability depends on deployment context, permissions, balances, and reachable preconditions.
    • Analyzer output did not include a source excerpt for this finding.
  • mediumAccess controlConfidence: unknownExploitability: unknownDetector: semgrep

    Use of tx.origin for authorization

    Affected location
    VegaTokenVulnerable.sol:46
    Detector / source
    semgrep · semgrep:vegasec-tx-origin-auth

    Evidence

    Use of tx.origin for authorization. tx.origin can be tricked via phishing contracts. Prefer msg.sender.

    Exploitability notes

    Analyzer did not report a confidence level; treat as unverified until reviewed.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • Analyzer did not report confidence; fail closed and treat the finding as unverified.
    • Static analysis did not prove exploitability; do not present this as an exploit claim.
    • Analyzer output did not include a source excerpt for this finding.
  • lowAccess controlConfidence: mediumExploitability: mediumDetector: slither

    VegaTokenVulnerable.unsafeReviewerLottery(address[]).picked (VegaTokenVulnerable.sol#28) lacks a zero-check on :

    Affected location
    VegaTokenVulnerable.sol:28 · picked
    Detector / source
    slither · slither:missing-zero-check

    Evidence

    VegaTokenVulnerable.unsafeReviewerLottery(address[]).picked (VegaTokenVulnerable.sol#28) lacks a zero-check on : - reviewer = picked (VegaTokenVulnerable.sol#32)

    Exploitability notes

    Medium-confidence detection; manual review required before relying on the finding either way.

    Remediation

    Gate privileged functions behind a tested role/ownership pattern (e.g. OpenZeppelin Ownable/AccessControl). Avoid tx.origin for authorization.

    Limitations

    • Medium analyzer confidence; manual review should confirm the code path and preconditions.
    • Exploitability depends on deployment context, permissions, balances, and reachable preconditions.
    • Analyzer output did not include a source excerpt for this finding.
  • informationalLogicConfidence: highExploitability: lowDetector: slither

    Version constraint ^0.8.20 contains known severe issues (https://solidity.readthedocs.io/en/latest/bugs.html)

    Affected location
    VegaTokenVulnerable.sol:2 · ^0.8.20
    Detector / source
    slither · slither:solc-version

    Evidence

    Version constraint ^0.8.20 contains known severe issues (https://solidity.readthedocs.io/en/latest/bugs.html) - VerbatimInvalidDeduplication - FullInlinerNonExpressionSplitArgumentEvaluationOrder - MissingSideEffectsOnSelectorAccess. It is used by: - ^0.8.20 (VegaTokenVulnerable.sol#2)

    Exploitability notes

    Informational signal; not directly exploitable without additional context.

    Remediation

    Re-derive the business rule by hand, write a property-based test that fails on the current path, and only then fix the implementation.

    Limitations

    • High analyzer confidence, still requires source-level review before promotion.
    • Direct exploitability is not established; keep the signal as a remediation/checklist item.
    • Analyzer output did not include a source excerpt for this finding.
  • informationalOtherConfidence: highExploitability: lowDetector: slither

    VegaTokenVulnerable.owner (VegaTokenVulnerable.sol#11) should be immutable

    Affected location
    VegaTokenVulnerable.sol:11 · owner
    Detector / source
    slither · slither:immutable-states

    Evidence

    VegaTokenVulnerable.owner (VegaTokenVulnerable.sol#11) should be immutable

    Exploitability notes

    Informational signal; not directly exploitable without additional context.

    Remediation

    Review the detector output against the source manually; if exploitable, design a fix and add a regression test before deploying.

    Limitations

    • High analyzer confidence, still requires source-level review before promotion.
    • Direct exploitability is not established; keep the signal as a remediation/checklist item.
    • Analyzer output did not include a source excerpt for this finding.
  • informationalUnchecked callConfidence: highExploitability: lowDetector: slither

    Low level call in VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42):

    Affected location
    VegaTokenVulnerable.sol:36-42 · withdrawReviewCredit
    Detector / source
    slither · slither:low-level-calls

    Evidence

    Low level call in VegaTokenVulnerable.withdrawReviewCredit(uint256) (VegaTokenVulnerable.sol#36-42): - (ok,None) = address(msg.sender).call{value: amount}() (VegaTokenVulnerable.sol#39)

    Exploitability notes

    Informational signal; not directly exploitable without additional context.

    Remediation

    Check the boolean return of low-level call/delegatecall/send and revert on failure with a descriptive reason string.

    Limitations

    • High analyzer confidence, still requires source-level review before promotion.
    • Direct exploitability is not established; keep the signal as a remediation/checklist item.
    • Analyzer output did not include a source excerpt for this finding.

Limitations

  • VegaSec runs static analysis only. Findings can include false positives, miss patterns the configured detectors do not cover, and never replace manual auditor judgement.
  • Scan ran slither and semgrep against a single sandboxed copy of the source. Multi-file dependency analysis beyond what slither resolves automatically is out of scope.
  • Some detectors emitted warnings (see `warnings`); their output was still ingested but should be read with extra care.
  • This report is not a certification, audit attestation, or financial advice.

Exploit reasoning

Reference-grade reasoning tied to evidence

  • 025d746d5dc01028d99ab6f083342b2f9b382eca5b913ce3121b6ef8656c9c07 · centralizationcritical

    Not determined

    Question: Is this finding exploitable in this specific contract?

    Reason: Analyzer confidence or exploitability is unknown; the narrative would overclaim without operator review.

  • 9e9cc0e9cfc74f54891b411b93c7449f2e172aa1da0a998e8e7b976ae04def6b · oracle-manipulationhigh
    Confidence: low

    Conditions point at the exploit but at least one precondition needs operator confirmation.

    Preconditions

    • Price or rate read from an on-chain source that can be moved in the same block by the caller
    • Decisions (mint amount, collateral check) derived from the manipulable price

    Action

    Attacker uses a flash loan or large swap to move the oracle price in the same block, triggers the price-dependent action, then reverts the price.

    Impact

    Mint of more shares than collateral warrants, undercollateralized borrow, or theft of liquidity by exploiting the manipulated rate.

    False-positive caveats

    • The oracle may be a time-weighted aggregator that resists single-block movement.
    • The function may compare against a snapshot taken on a separate block.

    Remediation checklist

    • Use a TWAP (time-weighted average price) over multiple blocks.
    • Cross-check at least two independent oracle sources before acting.
    • Reject prices that diverge from a recent average by more than a documented threshold.
  • e3c81b2bb05292a33bb195f5efdac2095dcb7378274eb86ee20c848b965cd13e · reentrancyhigh
    Confidence: low

    Conditions point at the exploit but at least one precondition needs operator confirmation.

    Preconditions

    • External call to an untrusted address before the contract's own state is updated
    • Contract or callback path reachable from the external caller
    • No nonReentrant guard or equivalent on the entry-point function

    Action

    Attacker contract receives control during the external call and re-enters the original function before the original state write completes, repeating the state-changing side effect.

    Impact

    Duplicated withdrawals, double-spend on accounting balances, or out-of-order updates to invariants the function relies on.

    False-positive caveats

    • The external call may be to a known, trusted contract whose behavior is provable.
    • The function may be guarded indirectly by a flag or external lock not visible to the analyzer.

    Remediation checklist

    • Apply checks-effects-interactions: update internal state before any external call.
    • Add a nonReentrant guard on the entry-point function.
    • Where feasible, pull-payment instead of push-payment on user balances.
  • 7d9b78b620334b47ab494454ec49b378daf245c24264ec5ee94fd81ae25ca8a0 · access-controlhigh
    Confidence: high

    Conditions for the exploit are met by the analyzer evidence; treat as likely until disproven by operator review.

    Preconditions

    • Sensitive function lacks an explicit caller restriction or relies on tx.origin
    • Caller controls the path that reaches the function

    Action

    Attacker calls the unprotected function from an attacker-controlled address, bypassing the intended restriction.

    Impact

    Unauthorized state mutation: ownership transfer, parameter changes, balance writes, or admin-only operations executed by a non-admin.

    False-positive caveats

    • The function may be intentionally callable by anyone (e.g. a public mint with a price guard).
    • Access control may live in a modifier the analyzer did not parse.

    Remediation checklist

    • Replace tx.origin checks with msg.sender against a role registry.
    • Add a `onlyRole(...)` or equivalent modifier from a vetted access control library.
    • Add unit tests that assert non-admin callers revert.
  • 8d94c22f92bfff04d2cec20f538c96db7e75d788478a91bc7ffd1cceea98d7a0 · access-controlmedium
    Confidence: low

    Conditions point at the exploit but at least one precondition needs operator confirmation.

    Preconditions

    • Sensitive function lacks an explicit caller restriction or relies on tx.origin
    • Caller controls the path that reaches the function

    Action

    Attacker calls the unprotected function from an attacker-controlled address, bypassing the intended restriction.

    Impact

    Unauthorized state mutation: ownership transfer, parameter changes, balance writes, or admin-only operations executed by a non-admin.

    False-positive caveats

    • The function may be intentionally callable by anyone (e.g. a public mint with a price guard).
    • Access control may live in a modifier the analyzer did not parse.

    Remediation checklist

    • Replace tx.origin checks with msg.sender against a role registry.
    • Add a `onlyRole(...)` or equivalent modifier from a vetted access control library.
    • Add unit tests that assert non-admin callers revert.
  • d3a46569a601f7bdcf08c52c1ce6a464f24ea0d6035d39fa18437b6c940bbc00 · access-controlmedium

    Not determined

    Question: Is this finding exploitable in this specific contract?

    Reason: Analyzer confidence or exploitability is unknown; the narrative would overclaim without operator review.

  • 4bf225cdc0c238c3ce62538e80d89916562e5eb0b7ed3c988eaf7bb73b29c755 · timestamp-dependencemedium

    Not determined

    Question: Is this finding exploitable in this specific contract?

    Reason: Analyzer confidence or exploitability is unknown; the narrative would overclaim without operator review.

  • 24612734790f3f6ca717a5032967841c609ae6046549b32eb41f771d29f905a2 · unchecked-callmedium

    Not determined

    Question: Is this finding exploitable in this specific contract?

    Reason: Analyzer confidence or exploitability is unknown; the narrative would overclaim without operator review.

  • b85d4c6b14a005cc536ebad1f32253f03512ff61c8e9c49a7cbf0952e914f482 · access-controlmedium
    Confidence: low

    Conditions point at the exploit but at least one precondition needs operator confirmation.

    Preconditions

    • Sensitive function lacks an explicit caller restriction or relies on tx.origin
    • Caller controls the path that reaches the function

    Action

    Attacker calls the unprotected function from an attacker-controlled address, bypassing the intended restriction.

    Impact

    Unauthorized state mutation: ownership transfer, parameter changes, balance writes, or admin-only operations executed by a non-admin.

    False-positive caveats

    • The function may be intentionally callable by anyone (e.g. a public mint with a price guard).
    • Access control may live in a modifier the analyzer did not parse.

    Remediation checklist

    • Replace tx.origin checks with msg.sender against a role registry.
    • Add a `onlyRole(...)` or equivalent modifier from a vetted access control library.
    • Add unit tests that assert non-admin callers revert.
  • 2b383bb32009d69610ac628adccd9eee4b96f1e5449289991c5149e18497756b · access-controlmedium

    Not determined

    Question: Is this finding exploitable in this specific contract?

    Reason: Analyzer confidence or exploitability is unknown; the narrative would overclaim without operator review.

Judgment

Triage of the findings, in reading order

Operator-reviewed

The ranking, the explanations and the remediation order below were produced by a language model from the findings this report already carries (model deepseek/deepseek-v4.1-flash) generated Sep 27, 2026, 3:59 AM. It is a triage layer, not a detection layer: every finding comes from the analyzers above.

Reviewed by a VegaSec operator on Sep 27, 2026, 3:59 AM. The review covers exactly the content shown here.

Ranking

  1. #1 · 7d9b78b620334b47ab494454ec49b378daf245c24264ec5ee94fd81ae25ca8a0 · access-controlhigh

    High confidence, concrete: retireFixture() selfdestructs the contract and is gated only by tx.origin, which a phishing contract can satisfy on the owner's behalf.

  2. #2 · 025d746d5dc01028d99ab6f083342b2f9b382eca5b913ce3121b6ef8656c9c07 · centralizationcritical

    Critical selfdestruct on the same line 47 as the suicidal finding; separate rule id, so read together with it rather than as a second path.

  3. #3 · e3c81b2bb05292a33bb195f5efdac2095dcb7378274eb86ee20c848b965cd13e · reentrancyhigh

    External value call to msg.sender before the balance decrement in withdrawReviewCredit, so the caller can re-enter and drain credited balances.

  4. #4 · 9e9cc0e9cfc74f54891b411b93c7449f2e172aa1da0a998e8e7b976ae04def6b · oracle-manipulationhigh

    unsafeReviewerLottery selects the reviewer from block.timestamp, which the block producer influences, making the stored reviewer predictable.

  5. #5 · 8d94c22f92bfff04d2cec20f538c96db7e75d788478a91bc7ffd1cceea98d7a0 · access-controlmedium

    mintForReview authorizes by tx.origin, so any contract the owner calls can mint to an arbitrary account.

  6. #6 · b85d4c6b14a005cc536ebad1f32253f03512ff61c8e9c49a7cbf0952e914f482 · access-controlmedium

    Second tx.origin gate, inside retireFixture; it is the authorization half of the destruction path and must be fixed with it.

  7. #7 · d3a46569a601f7bdcf08c52c1ce6a464f24ea0d6035d39fa18437b6c940bbc00 · access-controlmedium

    Semgrep duplicate of the tx.origin check at line 24; no new information beyond 8d94c22f.

  8. #8 · 2b383bb32009d69610ac628adccd9eee4b96f1e5449289991c5149e18497756b · access-controlmedium

    Semgrep duplicate of the tx.origin check at line 46 in retireFixture; same site as b85d4c6b.

  9. #9 · 4bf225cdc0c238c3ce62538e80d89916562e5eb0b7ed3c988eaf7bb73b29c755 · timestamp-dependencemedium

    Timestamp-as-randomness at line 31, duplicate of the weak-PRNG finding; keep as corroboration of the same defect.

  10. #10 · 24612734790f3f6ca717a5032967841c609ae6046549b32eb41f771d29f905a2 · unchecked-callmedium

    Flags the payout call target at line 39 as externally controlled; it is the mechanism behind the withdrawReviewCredit reentrancy.

  11. #11 · c95e896eba0879481c44b2f622b6ae140a700984244981b4404ff41210f6c2d3 · unchecked-callinformational

    Informational restatement of the same low-level call site already covered by e3c81b2b and 24612734.

  12. #12 · 0e297411d10f1e4a0ee9692c268471424d3d570524701f41a065fe7ac3d700e9 · access-controllow

    Low-severity zero-check note on the named return of unsafeReviewerLottery; the function does require non-empty input, so impact is limited.

  13. #13 · 5000022c8e54141e1f7c204b377ec3cddd18eb4f224def2ddcfff5238884b60b · logicinformational

    Informational compiler-range note on pragma ^0.8.20; no exploit path in this contract but worth pinning at build time.

  14. #14 · a2093d5f0250fce2f6c8bbc1c6c8c09b8ac4c01d16989600d7b91a130bb8aded · otherinformational

    Informational gas/style note that owner could be immutable; unrelated to the fund-loss and destruction paths above.

Why it matters

  • 7d9b78b620334b47ab494454ec49b378daf245c24264ec5ee94fd81ae25ca8a0

    retireFixture() destroys the contract and forwards the balance to owner. The only gate is tx.origin == owner, so a contract the owner interacts with can call retireFixture() while the owner's transaction is in flight; tx.origin still equals owner and the call succeeds. That ends the token contract and its state in one transaction, which no later fix can undo.

    Cited line · VegaTokenVulnerable.sol:47

    selfdestruct(payable(owner));
  • 025d746d5dc01028d99ab6f083342b2f9b382eca5b913ce3121b6ef8656c9c07

    Same line as the suicidal finding: the selfdestruct in retireFixture(). Reported as critical centralization because the destruction capability is tied to a single owner-controlled gate rather than a reviewable shutdown procedure. Whatever the intended lifecycle, once this call executes the deployed address is dead and balances at that address are no longer accessible through normal calls.

    Cited line · VegaTokenVulnerable.sol:47

    selfdestruct(payable(owner));
  • e3c81b2bb05292a33bb195f5efdac2095dcb7378274eb86ee20c848b965cd13e

    withdrawReviewCredit checks balances[msg.sender] >= amount, then sends value to msg.sender with a low-level call, and only afterwards subtracts from balances. During the call the caller's balance is still whole, so a receiving contract can call withdrawReviewCredit again and be paid repeatedly until the contract's ether is gone. The subtraction at line 41 happens after control returns.

    Cited line · VegaTokenVulnerable.sol:39

    (bool ok,) = payable(msg.sender).call{value: amount}("");
  • 9e9cc0e9cfc74f54891b411b93c7449f2e172aa1da0a998e8e7b976ae04def6b

    unsafeReviewerLottery picks a reviewer with block.timestamp modulo the array length. The block producer chooses that timestamp within its allowed drift, so the selected reviewer and the reviewer state variable written at line 32 are predictable and steerable. Anything that later trusts reviewer as an impartial selection inherits that manipulation.

    Cited line · VegaTokenVulnerable.sol:31

    picked = reviewers[block.timestamp % reviewers.length];
  • 8d94c22f92bfff04d2cec20f538c96db7e75d788478a91bc7ffd1cceea98d7a0

    mintForReview credits arbitrary balances after checking tx.origin == owner. Any contract the owner transacts with can invoke mintForReview during that transaction, because tx.origin remains the owner while msg.sender is the attacker's contract. The minted amount and recipient are both caller-supplied, so the caller chooses who gets credited.

    Cited line · VegaTokenVulnerable.sol:24

    require(tx.origin == owner, "origin owner only");

Remediation order

  1. Delete the selfdestruct from retireFixture and from the contract. If shutdown must exist, replace it with a pausable or withdraw-and-freeze path gated on msg.sender and reviewed before deployment.7d9b78b620334b47ab494454ec49b378daf245c24264ec5ee94fd81ae25ca8a0
  2. Close out the same line: confirm no other function reaches selfdestruct and re-scan after removal, since this rule and the suicidal rule both point at line 47.025d746d5dc01028d99ab6f083342b2f9b382eca5b913ce3121b6ef8656c9c07
  3. Replace tx.origin == owner with msg.sender == owner (or a role check) in retireFixture; if the function is removed, drop the check with it.b85d4c6b14a005cc536ebad1f32253f03512ff61c8e9c49a7cbf0952e914f482
  4. Same site as priority 3: after the gate is changed to msg.sender, re-run semgrep so the line-46 tx.origin hit clears as well.2b383bb32009d69610ac628adccd9eee4b96f1e5449289991c5149e18497756b
  5. Change mintForReview to authorize on msg.sender, or better to an explicit minter role, so a phishing contract cannot mint on the owner's behalf.8d94c22f92bfff04d2cec20f538c96db7e75d788478a91bc7ffd1cceea98d7a0
  6. Duplicate of priority 5 at line 24; verify the semgrep hit clears once msg.sender authorization is in place.d3a46569a601f7bdcf08c52c1ce6a464f24ea0d6035d39fa18437b6c940bbc00
  7. In withdrawReviewCredit, subtract balances[msg.sender] -= amount before the external call, or add a reentrancy guard, then keep the call and the ok check after effects.e3c81b2bb05292a33bb195f5efdac2095dcb7378274eb86ee20c848b965cd13e
  8. After the ordering fix, consider replacing the raw call with a pull-payment pattern or a limited-gas send so the recipient cannot re-enter on this path.24612734790f3f6ca717a5032967841c609ae6046549b32eb41f771d29f905a2
  9. Replace block.timestamp selection in unsafeReviewerLottery with a commit-reveal scheme or an external randomness beacon; storing a producer-influenced pick is not sufficient.9e9cc0e9cfc74f54891b411b93c7449f2e172aa1da0a998e8e7b976ae04def6b
  10. Same line 31 as priority 9; confirm the timestamp use is gone before re-scanning.4bf225cdc0c238c3ce62538e80d89916562e5eb0b7ed3c988eaf7bb73b29c755
  11. While editing unsafeReviewerLottery, add an explicit check on the chosen reviewer address before assigning it to reviewer.0e297411d10f1e4a0ee9692c268471424d3d570524701f41a065fe7ac3d700e9
  12. Pin the pragma to a specific compiler release and re-review against the linked known-issues list before deployment.5000022c8e54141e1f7c204b377ec3cddd18eb4f224def2ddcfff5238884b60b
  13. If owner is set once in the constructor, mark it immutable; otherwise leave as is and document why it is mutable.a2093d5f0250fce2f6c8bbc1c6c8c09b8ac4c01d16989600d7b91a130bb8aded

Uncertain

  • 025d746d5dc01028d99ab6f083342b2f9b382eca5b913ce3121b6ef8656c9c07
  • 8d94c22f92bfff04d2cec20f538c96db7e75d788478a91bc7ffd1cceea98d7a0
  • 9e9cc0e9cfc74f54891b411b93c7449f2e172aa1da0a998e8e7b976ae04def6b
  • 0e297411d10f1e4a0ee9692c268471424d3d570524701f41a065fe7ac3d700e9

Citation check

36 of 36 produced items were kept; 5quoted line(s) were checked against the report's own source; nothing was discarded.

What this section does not claim

  • It does not detect anything: the finding set is the input, never an output.
  • It is not a human review. A machine triage is not an audit, and this section is not evidence that the contract is free of defects.
  • An empty or short triage is not evidence of anything either: no finding set here means the analyzers raised nothing on this source.
  • Ordering is one reading of the report, not a verdict on exploitability.

Audit pipeline

Audit cockpit

Honest stage-by-stage view of what VegaSec ran against this contract. Automated stages do not replace a human operator review.

Passed: 3Partial: 5Blocked: 1
  1. Audit intake

    Passed320ms replay4:42 PM

    Normalize the submitted contract target and confirm it can enter the Solidity audit lane.

    Next: Acquire verified source or scan the pasted Solidity source.

  2. Source acquisition

    Passed1.4s replay4:42 PM

    Resolve pasted source or verified explorer source without treating malformed input as demo data.

    Next: Validate the analyzer toolchain against the acquired source.

  3. Toolchain validation

    Partial2.3s replay4:42 PM

    Confirm the configured static analyzers can run and produce parseable output.

    Next: Run static analysis and capture raw detector evidence.

    • One or more analyzers emitted warnings; report findings remain usable but should be read with limitations.
  4. Static analysis

    Partial14 findings8.8s replay4:42 PM

    Run static analyzers against the sandboxed source and collect raw detector evidence.

    Next: Normalize detector output into one deduped finding set.

    • One or more analyzers emitted warnings; report findings remain usable but should be read with limitations.
  5. Finding normalization

    Passed14 findings1.1s replay4:42 PM

    Dedupe detectors, classify vulnerability classes and compute severity-weighted risk.

    Next: Review exploitability conservatively against confidence and limitations.

  6. Exploitability review

    Partial14 findings1.9s replay4:42 PM

    Frame automated findings with conservative exploitability and limitation notes.

    Next: Draft remediation notes from the normalized evidence.

    • Automated exploitability review is conservative and does not replace a manual exploitability assessment.
  7. Remediation draft

    Partial14 findings1.5s replay4:42 PM

    Prepare remediation-oriented evidence for the shareable report artifact.

    Next: Attach an operator review artifact before claiming human review.

    • Remediation notes are generated from detector evidence and require project-specific engineering review.
  8. Operator review

    Blocked0ms replay4:42 PM

    Record whether a human operator artifact exists before buyer escalation.

    Next: Attach a real operator review artifact before claiming human review.

    • No operator review artifact is attached to this automated scan.
    • Human operator review has not been attached to this automated scan yet.
  9. Finalization

    Partial14 findings620ms replay4:42 PM

    Publish the automated report artifact or explain why the scan failed closed.

    Next: Share the report with explicit limitations and review state.

    Artifact: Automated Solidity report · solidity-2f746d702f766567

    • Final report is an automated static-analysis artifact; no operator review artifact is attached.

Report badges

Verified source

Badges describe the evidence carried by this report artifact, not financial value or future pricing. Operator-reviewed is never granted automatically.

How to read this report

Score, confidence, ranking: four moving parts, no hand-tuning.

Generated Sep 24, 2026, 4:42 PM · Solidity · VegaTokenVulnerable.sol

Score bands

  • 0-19

    AVOID-style triage · this report

    Multiple high or critical findings, low source confidence, or unverified context. Public action should pause until a Private Review or a re-scan with better evidence.

  • 20-49

    High caution

    Notable findings or thin evidence. Worth a Private Review before relying on the artifact.

  • 50-69

    Watch

    Mixed signals: some evidence, some gaps. Re-scan after material changes to the target.

  • 70-89

    Clear surface (automated)

    No critical or high findings via static analysis. Not a safety claim: human review is still recommended for diligence work.

  • 90-100

    Strong surface (automated)

    No critical or high findings, verified source, and high source confidence. Not a safety claim: human review is still recommended for diligence work.

Score
Deterministic 0-100 derived from severity counts. Same formula across every report. Lower means more risk.
Confidence
Separate axis. Measures how much VegaSec trusts the input data. Low confidence pushes the verdict toward UNKNOWN; the two are never collapsed into one number.
Ranking / percentile
Only meaningful within the eligible Trust Index set. Shown on /trust-index when the eligible pool is large enough. Today the public index is a controlled preview.

Private Review is separate

A paid Private Review does not raise the public score on this report. Public score reflects public-source evidence only. Private Review adds operator attestation and a confidential memo, never a score bump.

Export and share

  • · Shareable public URL: copy from the report actions panel.
  • · Markdown export: every export carries the report id, target, verdict, generated timestamp, and the public-data disclaimer.
  • · PDF export: not in this MVP. Tracked as a follow-up. The Markdown export contains the same content and prints cleanly to PDF from a browser today.

Source provenance

This report scanned a Solidity source provided directly by the submitter (paste). No on-chain address is associated.

Scanned source (2,036 chars)

// SPDX-License-Identifier: MIT
pragma solidity ^0.8.20;

/// @title VegaTokenVulnerable
/// @notice Fictional internal VegaSec self-audit fixture only. Not deployed, not sold, no financial promise.
/// @dev This intentionally unsafe contract exists so VegaSec can prove vulnerable-vs-fixed audit deltas.
contract VegaTokenVulnerable {
    string public constant FIXTURE_SCOPE = "VegaSec fictional internal self-audit fixture";

    mapping(address => uint256) public balances;
    address public owner;
    address public reviewer = 0x1111111111111111111111111111111111111111;
    uint256 public reviewRound;

    constructor() payable {
        owner = msg.sender;
        balances[msg.sender] = 1_000_000;
    }

    receive() external payable {}

    function mintForReview(address account, uint256 amount) external {
        // Intentional issue: tx.origin authorization can be phished through another contract.
        require(tx.origin == owner, "origin owner only");
        balances[account] += amount;
    }

    function unsafeReviewerLottery(address[] calldata reviewers) external returns (address picked) {
        require(reviewers.length > 0, "reviewers required");
        // Intentional issue: block.timestamp is miner/validator-influenced and unsafe as randomness.
        picked = reviewers[block.timestamp % reviewers.length];
        reviewer = picked;
        reviewRound += 1;
    }

    function withdrawReviewCredit(uint256 amount) external {
        require(balances[msg.sender] >= amount, "insufficient credits");
        // Intentional issue: external low-level call before effects enables reentrancy-style drains.
        (bool ok,) = payable(msg.sender).call{value: amount}("");
        require(ok, "transfer failed");
        balances[msg.sender] -= amount;
    }

    function retireFixture() external {
        // Intentional issue: tx.origin gate plus selfdestruct is unsafe and should never ship.
        require(tx.origin == owner, "origin owner only");
        selfdestruct(payable(owner));
    }
}

Public-data disclaimer

Solidity scans are triage based on slither + semgrep static analysis. Output is not a certification, audit attestation or financial advice. Findings can include false positives and may miss patterns the configured detectors do not cover.