fix(security): Fix 31 security issues in axios, tar, hono

[aikido-autofix]

Upgrade axios, tar, and hono to fix critical SSRF/proxy bypass, prototype pollution MITM attacks, and header injection vulnerabilities.

✅ 31 CVEs resolved by this upgrade, including 1 critical 🚨 CVE

This PR will resolve the following CVEs:

Issue	Severity	Description
CVE-2025-62718	🚨 CRITICAL	[axios] Improper hostname normalization in NO_PROXY rule checking allows requests to loopback addresses (localhost., [::1]) to bypass proxy protections, enabling proxy bypass and potential SSRF attacks against internal services. This vulnerability permits attackers to reach sensitive services despite configured NO_PROXY protections.
CVE-2026-44494	HIGH	[axios] Prototype pollution vulnerability allows escalation to Man-in-the-Middle attacks by injecting a malicious proxy into the prototype chain, enabling interception and modification of all HTTP traffic including authentication credentials. An attacker can read and alter all application HTTP communications through a compromised dependency.
AIKIDO-2026-10741	HIGH	[axios] An HTTP client vulnerability allows prototype pollution through loose object merging, sensitive data exposure via error serialization, and improper proxy/socket handling that could lead to information disclosure or unauthorized access.
CVE-2026-40175	HIGH	[axios] A prototype pollution vulnerability in a third-party dependency can be exploited to inject unsanitized header values into outbound HTTP requests. This allows attackers to manipulate request headers, potentially leading to request smuggling, cache poisoning, or other header-based attacks.
AIKIDO-2026-10823	HIGH	[axios] HTTP adapter incorrectly routes HTTPS traffic through HTTP proxies in cleartext instead of using CONNECT tunnels, allowing proxies to observe sensitive request metadata and payloads that should be encrypted.
AIKIDO-2026-291630	HIGH	[axios] Cross-origin redirects leak custom credential headers like X-API-Key and AWS tokens to unintended hosts, allowing attackers to steal sensitive authentication data. This information disclosure vulnerability affects shared environments where secret headers are set by default.
AIKIDO-2026-10509	HIGH	[axios] Prototype pollution vulnerability allows attackers to inject malicious headers into requests through unsafe FormData detection and header merging, potentially enabling authorization bypass or request manipulation.
GHSA-42h9-826w-cgv3	MEDIUM	[axios] Uncontrolled recursion in formDataToJSON when processing deeply nested FormData field names causes stack overflow, resulting in denial of service for applications converting untrusted FormData to JSON.
GHSA-pmv8-rq9r-6j72	MEDIUM	[axios] Uncontrolled recursion in formDataToJSON allows attackers to crash the process by submitting FormData with deeply nested bracket-delimited field names, causing stack exhaustion and denial of service.
GHSA-7q8q-rj6j-mhjq	MEDIUM	[axios] Prototype pollution in nested request options allows inherited Object.prototype properties to inject malicious Authorization headers and manipulate query strings when applications pass placeholder objects like auth: {} or paramsSerializer: {}. This enables credential exfiltration and request tampering when another component has already polluted the prototype chain.
GHSA-jqh4-m9w3-8hp9	MEDIUM	[axios] The fetch adapter fails to enforce maxBodyLength limits on WHATWG ReadableStream request bodies with unknown sizes, allowing attackers to bypass upload size restrictions and cause unexpected network egress or resource exhaustion.
GHSA-mmx7-hfxf-jppx	MEDIUM	[axios] Prototype pollution read-side gadgets in axios allow attackers who pollute Object.prototype to hijack requests by injecting malicious data, proxy settings, or parameter serializers into bodyless methods and low-level adapter calls. This can result in credential theft, request body injection, traffic redirection, or URL manipulation depending on the attack vector.
AIKIDO-2026-38469	MEDIUM	[axios] Malformed HTTP/HTTPS URLs without // (e.g., https:internal.example) were silently normalized instead of rejected, allowing attackers to bypass URL allowlists or WAF checks and reach unintended hosts. The fix now throws an error for invalid scheme URLs before normalization.
AIKIDO-2026-10820	MEDIUM	[axios] Request size limits were not enforced in the fetch adapter, allowing responses and bodies to exceed configured limits, potentially causing denial of service or memory exhaustion. The vulnerability has been fixed to properly reject oversized responses and bodies.
AIKIDO-2026-10822	LOW	[axios] A prototype pollution vulnerability in the formDataToJSON helper allows attackers to pollute Object.prototype through specially crafted multipart field names, potentially enabling remote code execution or other malicious actions.
GHSA-mwf2-3pr3-8698	LOW	[axios] Streamed request bodies bypass maxBodyLength enforcement when sent over HTTP/2, allowing attackers to transmit arbitrarily large payloads and exhaust resources or bypass configured egress limits. This affects Node.js applications forwarding untrusted streams to HTTP/2 endpoints.
AIKIDO-2026-10819	LOW	[axios] A regular expression vulnerability in browser cookie parsing allowed metacharacters in cookie names to cause incorrect matching or excessive CPU consumption. The vulnerability was fixed by switching to literal string comparison instead of regex-based matching.
AIKIDO-2026-732519	HIGH	[tar] A crafted tar archive with a negative base-256 encoded size can cause the library to loop infinitely on the same header during archive scanning, resulting in a denial of service that pins the worker process.
AIKIDO-2026-986821	HIGH	[tar] A decompression bomb vulnerability allows attackers to exhaust disk storage by crafting TAR archives with highly compressible data claiming massive file sizes, causing denial of service through uncontrolled resource consumption during extraction.
CVE-2026-29786	MEDIUM	[tar] A vulnerability allows attackers to create hardlinks pointing outside the extraction directory using drive-relative paths, enabling arbitrary file overwrite during extraction. This could lead to remote code execution or data tampering depending on the target files.
CVE-2026-23745	MEDIUM	[tar] A vulnerability in tar extraction fails to sanitize link paths in hardlink and symbolic link entries, allowing malicious archives to bypass path restrictions and overwrite arbitrary files or create symlink poisoning attacks.
AIKIDO-2026-11058	MEDIUM	[tar] A parser differential vulnerability allows attackers to craft tar archives where PAX extended headers override sizes of metadata entries, causing tar to read different members than other tools and enabling malicious files to be hidden from security scanners while still being extracted downstream.
CVE-2026-31802	MEDIUM	[tar] A vulnerability allows attackers to create symlinks pointing outside the extraction directory using drive-relative paths, enabling arbitrary file overwrite during extraction. This could lead to remote code execution or data tampering depending on the target files.
AIKIDO-2026-568044	MEDIUM	[tar] A parsing vulnerability allows crafted PAX extended headers with embedded NUL bytes in paths to bypass validation and cause uncaught exceptions during archive extraction, enabling denial of service and validator bypass attacks.
AIKIDO-2026-401672	MEDIUM	[tar] A PAX header parser vulnerability coerces path values to numbers, causing an uncaught TypeError when downstream code calls .split('/') on the numeric value, crashing the Node.js process during archive extraction.
AIKIDO-2026-11187	HIGH	[hono] CORS middleware with credentials enabled and wildcard origin incorrectly reflects the request Origin header instead of rejecting it, allowing any third-party page to make credentialed cross-origin requests and read cookie-authenticated API responses.
AIKIDO-2026-571543	MEDIUM	[hono] The body-limit middleware on AWS Lambda adapters can be bypassed by declaring a small Content-Length header while sending a larger actual body, allowing oversized payloads to pass validation checks. This vulnerability enables denial-of-service attacks and potential remote code execution through request body processing.
AIKIDO-2026-11189	MEDIUM	[hono] A path traversal vulnerability in the serve-static middleware allows attackers to bypass authentication guards on Windows by using encoded backslashes in URLs to access protected static files. The vulnerability enables unauthenticated file disclosure without escaping the configured root directory.
GHSA-26pp-8wgv-hjvm	MEDIUM	[hono] Cookie names are not validated in setCookie(), serialize(), or serializeSigned(), allowing invalid characters that can cause malformed Set-Cookie headers and runtime errors when processing untrusted cookie names.
AIKIDO-2026-943075	MEDIUM	[hono] The AWS Lambda adapter incorrectly merges multiple Set-Cookie headers into a single comma-separated value, causing clients to misparsed cookies and potentially drop session or CSRF tokens, breaking authentication. This is fixed by emitting Set-Cookie as an array instead.
AIKIDO-2026-362336	MEDIUM	[hono] A Lambda@Edge adapter vulnerability causes multi-value HTTP headers to be overwritten instead of preserved, resulting in incomplete header chains reaching application middleware and potentially bypassing access control or auditing mechanisms that depend on complete header data.

[github-actions]

🚀 Preview deployed!

Worker: lintent-pr-66
URL: https://lintent-pr-66.li-fi374.workers.dev