CVE-2026-77261 — mcp-atlassian

Summary

_make_ssrf_safe_hook() blocks HTTP redirects to private/internal IPs by validating the Location header before the client follows a 3xx response. The problem is that this hook is only attached in one of three authentication branches — the header-PAT path. Basic auth and OAuth branches skip it entirely, so if the connected Atlassian server returns a redirect to something like http://169.254.169.254/, the requests session follows it without complaint.

This is an incomplete fix for GHSA-7r34-79r5-rcc9. The hook works fine when it's there — it just isn't there for most production auth configurations.

Details

In src/mcp_atlassian/servers/dependencies.py, three branches construct a fetcher and call _create_and_validate(). Only Branch 1 passes attach_ssrf_hook=True:

# Branch 1 (header PAT) — hook attached
return _create_and_validate(request, spec, header_config, "header_pat",
                             attach_ssrf_hook=True)

# Branch 2 (basic auth) — hook missing
return _create_and_validate(request, spec, user_config, "basic",
                             user_email=user_email)

# Branch 3 (OAuth/PAT) — hook missing
return _create_and_validate(request, spec, user_config, "oauth_pat",
                             user_email=user_email)

attach_ssrf_hook defaults to False, so branches 2 and 3 silently skip the protection. The hook itself (_make_ssrf_safe_hook) is straightforward — it checks response.is_redirect, grabs the Location header, and calls validate_url_for_ssrf() to reject private IPs. It works correctly when present.

Typical attack flow:

1. Attacker controls or compromises an Atlassian instance (Cloud or Server)
2. MCP server connects using basic auth or OAuth credentials (most production setups)
3. Atlassian returns 302 Location: http://169.254.169.254/latest/meta-data/iam/security-credentials/
4. The unprotected session follows the redirect
5. AWS IAM credentials (or other internal service data) are returned to the attacker

PoC

Tested on commit d8bc786 (v0.21.1). No real credentials needed.

from unittest.mock import MagicMock
import requests

from mcp_atlassian.servers.dependencies import _make_ssrf_safe_hook
from mcp_atlassian.utils.urls import validate_url_for_ssrf
from mcp_atlassian.jira import JiraFetcher
from mcp_atlassian.jira.config import JiraConfig

config = JiraConfig(
    url="https://attacker.atlassian.net",
    auth_type="basic",
    username="[email protected]",
    api_token="victim-token",
)
fetcher = JiraFetcher(config=config)
session = fetcher.jira._session

hooks = session.hooks.get("response", [])
print("hooks on basic-auth session:", [h.__name__ for h in hooks] or "none")

fake_redirect = MagicMock(spec=requests.Response)
fake_redirect.is_redirect = True
fake_redirect.headers = {"Location": "http://169.254.169.254/latest/meta-data/"}

blocked = False
for h in hooks:
    try:
        h(fake_redirect)
    except ValueError as e:
        blocked = True
        print("blocked:", e)

if not blocked:
    print("redirect to 169.254.169.254 not blocked on basic-auth session")

# show header-PAT branch does block it
hook = _make_ssrf_safe_hook(validate_url_for_ssrf)
try:
    hook(fake_redirect)
except ValueError as e:
    print("header-PAT branch blocks:", e)

Output:

$ uv run python3 /tmp/test.py
hooks on basic-auth session: none
redirect to 169.254.169.254 not blocked on basic-auth session
header-PAT branch blocks: Redirect blocked (SSRF): Blocked IP address: 169.254.169.254 (non-global)

Impact

Basic auth and OAuth cover most production Atlassian Cloud deployments, so this affects the majority of HTTP-mode multi-user setups. An attacker with control over the Atlassian server can redirect MCP server requests to internal infrastructure — cloud metadata endpoints, internal Kubernetes API, databases, or any service reachable from the MCP server's network.

The fix is one line per affected branch: pass attach_ssrf_hook=True to _create_and_validate() in branches 2 and 3, the same way branch 1 already does.