Vulnerability Intelligence

Public vulnerability intelligence, dependency exposure and research context for npm packages.

🌐

axios

Promise-based HTTP client for browser and Node.js security research

Current tracked
Current tracked 1.18.0
Fixed 1.18.0+
Affected <1.8.2; 1.13.0-1.13.1; <1.15.0; 1.15.0; <1.16.0; malicious: 1.14.1, 0.30.4

Axios is a widely used promise-based HTTP client for browsers and Node.js. It is commonly placed on authentication, API integration, redirect, proxy and service-to-service boundaries. Its current security relevance includes SSRF and credential leakage via URL resolution behavior, proxy and NO_PROXY bypasses, HTTP/2 denial-of-service conditions, header handling risks, and a confirmed 2026 npm supply-chain compromise involving malicious axios versions. Production usage should prefer axios 1.18.0 or later and explicitly block malicious versions 1.14.1 and 0.30.4.

Category HTTP Client
Ecosystem npm / JavaScript
Common usage API clients, frontend requests, backend integrations, authentication flows, redirects, headers, JSON payloads and HTTP automation.
Risk model High-risk dependency when used with SSRF-sensitive flows, proxy rules, redirects, custom headers, credentials or untrusted URL construction.
Why it is widely used
Simple and consistent API compared to native HTTP implementations.
Works across browsers and Node.js with a unified programming model.
Provides built-in handling for JSON payloads, headers, redirects and timeouts.
Supports interceptors for authentication, logging, observability and request manipulation.
Frequently embedded in SDKs, cloud integrations, payment platforms and third-party connectors.
Commonly used in service-to-service architectures and distributed systems.
Acts as a trust-boundary component between user-controlled input and external services.
Frequently appears in SSRF-sensitive request flows and proxy-controlled environments.
Risk score 92
Known issues 17
Exploit maturity Public advisories, confirmed supply-chain compromise, reproducible PoCs and active operational relev
Vulnerability burndown
Mar 2025 May 2026
Critical High Medium
MTTR critical severity
7 days
No data
Library risk age
277 days
92% lower than last month
Total vulnerabilities
17 Vulnerabilities
Critical 2 High 7 Medium 8 Low 0
Severity Vulnerability name Library Surface Status Published date SLA Tags Actions
Critical Cross-site Request Forgery axios@0.22.0 HTTP client New May 28, 2025 7 days CSRF
Critical Malicious npm Package Versions Delivered Cross-Platform RAT axios@1.14.1; axios@0.30.4 Supply chain / npm package publication Remediated 2026-03-31 Remove malicious versions; rotate credentials; rebuild from clean environment Supply Chain
High Regular Expression Denial of Service axios@0.22.0 Regex New May 16, 2025 20 days ReDoS
High Prototype Pollution (mergeConfig) axios <0.31.1 Config handling New Jun 2025 30 days Prototype Pollution
High NO_PROXY Bypass via IPv4-Mapped IPv6 Addresses axios >=1.0.0 <1.16.0; axios <=0.31.1 Proxy / NO_PROXY handling Published 2026-05-29 Upgrade to 1.16.0 or 0.32.0 SSRF
High Incomplete Fix for NO_PROXY Bypass via 127.0.0.0/8 Loopback Subnet axios 1.15.0 Proxy / NO_PROXY handling Published 2026-05-05 Upgrade beyond 1.15.0 SSRF
High Uncontrolled Recursion (toFormData) Validated PoC axios <0.31.1 Serializer / multipart form-data conversion Published 2025-06-01 Upgrade to 0.31.1 DoS
High HTTP Response Splitting via Headers Validated PoC axios <0.31.0 Headers / CRLF propagation Published 2025-06-01 Upgrade to 0.31.0 Header Injection
High SSRF and Credential Leakage via Absolute URL Override axios <1.8.2 Request URL resolution Published 2025-03-06 Upgrade to 1.8.2 SSRF
Medium NO_PROXY Hostname Normalization Bypass Leading to SSRF axios <1.15.0; axios <0.31.0 Proxy / NO_PROXY handling Published 2026-04-09 Upgrade to 1.15.0 or 0.31.0 SSRF
Medium Cloud Metadata Exfiltration via Header Injection Chain axios >=1.0.0 <1.15.0; axios <0.31.0 Headers / proxy / metadata service Published 2026-04-09 Upgrade to 1.15.0 or 0.31.0 Header Injection
Medium HTTP/2 Session Cleanup Denial of Service axios >=1.13.0 <1.13.2 HTTP/2 adapter Published 2026-04-08 Upgrade to 1.13.2 DoS
Medium XSRF Token Leakage via Config Manipulation Validated PoC axios <0.31.1 HTTP headers / XSRF token propagation Published 2025-06-01 Upgrade to 0.31.1 Data Leak
Medium Bypass of maxContentLength (large response) Validated PoC axios <0.31.1 HTTP adapter / stream response handling Published 2025-06-01 Upgrade to 0.31.1 DoS
Medium Bypass of maxBodyLength (upload) Validated PoC axios <0.31.1 HTTP adapter / streamed upload handling Published 2025-06-01 Upgrade to 0.31.1 DoS
Medium Improper Encoding (NUL byte injection) Validated PoC axios <0.31.1 URL params / AxiosURLSearchParams serialization Published 2025-06-01 Upgrade to 0.31.1 Encoding
Medium Server-side Request Forgery (SSRF) Validated PoC axios <0.30.0 Request handling / user-controlled URL fetching Published 2025-05-01 Upgrade to 0.30.0 SSRF
No vulnerability records match the selected filter.
FikreSekhel Research

Research Notes

Behavioral findings, exploitability observations and operational dependency intelligence produced by FikreSekhel for this library.

AXIOS-RN-2026-001

HTTP/2 Session Cache Fragmentation via Non-Canonical Option Equivalence

Behavioral analysis of transport option structural equality causing HTTP/2 session pool fragmentation under dynamically generated semantically equivalent configurations.

Observed Low Confirmed
Research identified that Axios HTTP/2 session reuse logic relies on structural equality of http2Options rather than transport semantic equivalence. Dynamically generated option objects containing irrelevant undefined-valued keys can fragment the session cache, resulting in unnecessary HTTP/2 session creation and degraded pooling efficiency. The Axios maintainers reviewed and classified the behavior as intended API usage rather than a security advisory candidate.
Surface HTTP/2 Transport Session Pooling
Primitive Session Cache Fragmentation
Tested versions Axios repository state (May 2026)
Observed behavior

When requests target the same authority but supply dynamically varying http2Options objects containing irrelevant undefined-valued keys, Axios allocates distinct HTTP/2 sessions rather than reusing existing pooled sessions. This produces linear session growth, increased socket handle count, and elevated memory consumption under repeated request execution.

Security implication

No confidentiality, integrity, or isolation boundary violation was observed. No cross-session leakage, authority confusion, memory corruption, authentication bypass, or remote code execution primitive was identified. The behavior is best classified as an operational resource-efficiency caveat relevant to application-layer transport configuration normalization.

Mitigation

Applications deriving http2Options dynamically should normalize transport configuration objects prior to Axios invocation. Implement deterministic allowlisting, canonical ordering, and elimination of semantically irrelevant undefined-valued properties before request dispatch. Multi-tenant transport wrappers should avoid exposing raw low-level HTTP/2 transport options directly to untrusted request context.

Observed before mitigation
[CONTROL] i=250 rss=53MB heap=13MB handles=3
Active sessions: 1
Observed after mitigation
[FRAGMENTED] i=250 rss=68MB heap=17MB handles=254
Active sessions: ~300
View PoC code
const client = axios.create({
  baseURL: "http://localhost:3002",
  httpVersion: 2,
  http2Options: {
    sessionTimeout: 30000,
    ["irrelevant_" + i]: undefined
  }
});

for (let i = 0; i < 300; i++) {
  await client.get("/");
}
View mitigation code
const stableClient = axios.create({
  baseURL: "http://localhost:3002",
  httpVersion: 2,
  http2Options: {
    sessionTimeout: 30000
  }
});
View FRES detection heuristic
Flag runtime patterns where semantically equivalent transport configuration objects are reconstructed per request with varying shape but identical transport semantics.
Need private intelligence for your codebase? Request deeper analysis, exploitability review and dependency risk mapping from the Fikresekhel consulting team.