CVE-2026-41645 is a medium-severity (CVSS 5.3) Code Injection vulnerability in github.com/projectdiscovery/nuclei/v3. A fix is available for github.com/projectdiscovery/nuclei/v3 — see the affected versions and patch details below.
Nuclei: Environment variable disclosure via Response-Derived DSL Expressions
Exploitation Status
No confirmed exploitation observed yet
- CISA’s own triage has not observed active exploitation or public proof-of-concept code for this CVE as of its last assessment.
Exploitation and automatability from CISA’s SSVC triage for CVE-2026-41645.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
CVE-2026-41645 plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 377,636 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
github.com/projectdiscovery/nuclei/v3Real-time download stats are indexed for npm and PyPI packages. This vulnerability affects Go packages — download data is not available via public APIs for these ecosystems.
Description
A vulnerability in Nuclei's expression evaluation engine makes it possible for a malicious target server to inject and execute supported DSL expressions. This happens when HTTP response data containing helper/function syntax gets reused by multi-step templates. If the -env-vars / -ev option is explicitly enabled, this can expose host environment variables. That option is off by default, so standard configurations are not affected by the information disclosure risk.
Affected Component
The issue lives in expressions.Evaluate() at pkg/protocols/common/expressions/ and in the unresolved-variable validation path (hasLiteralsOnly()).
Description
expressions.Evaluate() replaces placeholders first, then scans the substituted output for expressions. Because of this two-pass approach, response-derived values (including extractor output and response body content) can be reinterpreted as DSL/helper syntax on the second pass.
When -env-vars (-ev) is enabled, environment variables get merged into the template variable map. A malicious target can return response data containing expressions like {{env_var_name}} which, when reused in a subsequent template request, resolve to actual environment variable values. This can expose sensitive host data like API keys, credentials, and tokens.
Without -ev enabled (the default), injected DSL expressions may still trigger helper functions such as {{md5("test")}}, but this has no meaningful security impact beyond unexpected behavior.
There is also a separate issue in hasLiteralsOnly(): it was evaluating helper expressions while deciding whether {{...}} contained unresolved variables, which caused validation logic to run side-effectful helpers even when the final request kept the value as a literal.
[!NOTE] The
-env-vars/-evoption is off by default. Users who have not explicitly turned it on are not affected by the information disclosure aspect of this vulnerability.
Affected Users
- CLI users running multi-step templates (with extractors or flow-based request chaining) that reuse response-derived values against untrusted or attacker-controlled targets, with the
-evflag enabled. - SDK users who have integrated Nuclei into platforms where
EnvironmentVariablesis set totrueand scan targets are not fully trusted.
Patches
- The vulnerability is fixed in Nuclei v3.8.0. Upgrading to this version is strongly recommended.
- Relevant fix references: #7221, #7321.
Mitigation
Upgrade to Nuclei v3.8.0. The updated evaluation logic now collects expressions from the original template text before placeholder substitution and only evaluates those template-authored expressions.
If you have -ev enabled, disable it when scanning untrusted targets to avoid environment variable disclosure.
Workarounds
If upgrading is not an option right now, make sure -env-vars / -ev is not enabled when running multi-step templates against untrusted targets.
Acknowledgments
Nuclei thanks @gnuletik for reporting this issue through responsible disclosure via [email protected]
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐹Go | github.com/projectdiscovery/nuclei/v3 | ≥ 3.0.0&&< 3.8.0 | 3.8.0go get github.com/projectdiscovery/nuclei/v3@v3.8.0 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for github.com/projectdiscovery/nuclei/v3, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update github.com/projectdiscovery/nuclei/v3 to 3.8.0 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2026-41645 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 Security's impact-aware SCA analyses which vulnerable code paths your application actually calls, so a match like CVE-2026-41645 can be triaged on real exposure rather than presence alone.
Tailored to CVE-2026-41645. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is CVE-2026-41645 in your dependencies?
O3 Security finds CVE-2026-41645 across Go dependencies, including transitive ones, and its impact-aware SCA ranks findings by whether your code actually calls the vulnerable path.