- Top reward
- $3,750
- Top severity paid
- High
- Assets in scope
- 11
- Public CVEs
- 60
- Last change
- No changes yet
At a glance
What to know before you start testing.
- Pays up to $3,750
- for high bugs, from $625.
- Report on Sorona
- Submit through the program page.
- Partial safe harbor
- Some legal protection; read the terms.
- Replies within 10 days
- As the policy promises.
- Publish after a fix
- Coordinate the timing with the company.
Rewards
What Sorona lists as the payout for each severity.
| Severity | Relative size | Reward |
|---|---|---|
| High |
|
$3,750 |
| Medium |
|
$938 |
| Low |
|
$625 |
- Response
-
Within 10 days
10 business days First response
- Safe harbor
-
Partial
Sorona will not pursue legal action against you for good-faith research submitted to us in line with this policy. We cannot speak for the upstream project, its vendor, or any third party, and we do not grant authorization on their behalf.
Summarised with AI from the program policy on Sorona, read 12 Nov 2026, and checked word for word against it. Read the original
Scope
11 assets you may test, 6 excluded.
Web 1
- https://workers.cloudflare.com
Products and areas 10
- Durable Objects in full: runtime semantics, the storage API, alarms, and the isolation between clients and sessions talking to the same object
- The binding layer that glues V8 to the C++ runtime, where a type or lifetime mistake crosses the JavaScript boundary
- The bindings workerd implements for a Worker to reach other services, including service bindings, queues and the KV, R2 and D1 interfaces the runtime exposes
- The inspector and devtools endpoint workerd opens, and the local development paths wrangler dev drives through it
- The Node.js compatibility shims exposed under nodejs_compat, where a shim can hand back more capability than the real API would
- The whole Workers runtime API surface: fetch, Request/Response, Headers, streams, WebSocket, the Cache API, WebCrypto, URL and text encoding, timers and everything else a Worker can call
- The workerd server binary: process startup, socket and service wiring, and the Cap'n Proto configuration and control plane an embedder feeds untrusted config into
- V8 isolate lifecycle management, isolate boundaries, and any state that survives a request or crosses between tenants sharing a machine
- WebAssembly hosting inside a Worker, including the boundary between guest module and host runtime
- workerd is one of the few Cloudflare repositories the program names on its own. The live scope table carries github.com/cloudflare/workerd as a SOURCE_CODE asset, bounty-eligible at a maximum severity of critical, with no binding to the reduced open-source row, so the default ladder up to $10,000 applies. The declaration is the repository itself rather than a hosted wrapper around it, and it covers the runtime as a whole
Out of scope (6)
- Billing and plan-enforcement issues, which the policy caps at Low severity and a $250 maximum payout
- Findings that require a maintainer to approve a GitHub Actions workflow before they trigger
- Local-only wrangler dev reproductions with no demonstrated production reachability
- Reports with no working proof of concept, which are not bounty-eligible at any severity
- The repositories Cloudflare marks explicitly not eligible for bounty: cloudflare/vinext-private, moq-rs, privacy-gateway-server-go, saffron and recapn
- V8 itself, which is Chrome VRP territory and already sits on the sorona.xyz paid rail as its own program
Rules
What the policy asks of you, and the bug types it won't reward, quoted from Sorona's policy. "We" in these lines means Sorona, not this site.
While testing
-
Submit through sorona.xyz first. We validate, coordinate with the program's security team, and add the payout uplift on top of whatever it awards.
-
A working proof of concept is required: no PoC, no bounty, and source-code-only analysis is Informational regardless of the severity you assign it
-
A local workerd build or wrangler dev reproduction is graded Informational unless you show the same code path is reachable in production
-
Cloudflare runs its own AI-assisted discovery and closes matching external reports as duplicates, so file early rather than polishing
-
Billing and plan-enforcement issues are capped at Low severity and $250 maximum payout however you frame them
-
The generic github.com/cloudflare catch-all scope pays a lower $250 to $3,500 band; the figures below apply because workerd is separately named
Not accepted
-
V8 itself, which is Chrome VRP territory and already sits on the sorona.xyz paid rail as its own program
-
Local-only wrangler dev reproductions with no demonstrated production reachability
-
Billing and plan-enforcement issues, which the policy caps at Low severity and a $250 maximum payout
-
Findings that require a maintainer to approve a GitHub Actions workflow before they trigger
-
Reports with no working proof of concept, which are not bounty-eligible at any severity
Summarised with AI from the program policy on Sorona, read 12 Nov 2026, and checked word for word against it. Read the original
Changes
Tracked since 11 Nov 2026. Last checked 1 hour ago.
No changes yet. When the rewards or scope move, you will see it here and in the RSS feed.
Public CVEs
Published vulnerabilities in this vendor's products, from the NVD.
| CVE | Summary | Severity | Published |
|---|---|---|---|
| CVE-2026-12523 | Summary Cloudflare quiche's HTTP/3 layer was discovered to be vulnerable to resource exhaustion (i.e., memory) by means of specially crafted HTTP/3 frames. Impact HTTP/3 defines multiple frame types to support HTTP message exchanges and connection management. Each frame has a length and a payload whose length depends on the frame type. quiche was found to be vulnerable when parsing some frame type | High 7.5 | 14 Jul 2026 |
| CVE-2026-2833 | An HTTP request smuggling vulnerability (CWE-444) was found in Pingora's handling of HTTP/1.1 connection upgrades. The issue occurs when a Pingora proxy reads a request containing an Upgrade header, causing the proxy to pass through the rest of the bytes on the connection to a backend before the backend has accepted the upgrade. An attacker can thus directly forward a malicious payload after a req | Critical 9.3 | 5 Mar 2026 |
| CVE-2026-2835 | An HTTP Request Smuggling vulnerability (CWE-444) has been found in Pingora's parsing of HTTP/1.0 and Transfer-Encoding requests. The issue occurs due to improperly allowing HTTP/1.0 request bodies to be close-delimited and incorrect handling of multiple Transfer-Encoding values, allowing attackers to send HTTP/1.0 requests in a way that would desync Pingora’s request framing from backend servers’ | Critical 9.3 | 5 Mar 2026 |
| CVE-2026-2836 | A cache poisoning vulnerability has been found in the Pingora HTTP proxy framework’s default cache key construction. The issue occurs because the default HTTP cache key implementation generates cache keys using only the URI path, excluding critical factors such as the host header (authority). Operators relying on the default are vulnerable to cache poisoning, and cross-origin responses may be impr | High 8.4 | 5 Mar 2026 |
| CVE-2026-1229 | The CombinedMult function in the CIRCL ecc/p384 package (secp384r1 curve) produces an incorrect value for specific inputs. The issue is fixed by using complete addition formulas. ECDH and ECDSA signing relying on this curve are not affected. The bug was fixed in v1.6.3 https://github.com/cloudflare/circl/releases/tag/v1.6.3 . | Low 2.9 | 24 Feb 2026 |
| CVE-2026-0933 | SummaryA command injection vulnerability (CWE-78) has been found to exist in the `wrangler pages deploy` command. The issue occurs because the `--commit-hash` parameter is passed directly to a shell command without proper validation or sanitization, allowing an attacker with control of `--commit-hash` to execute arbitrary commands on the system running Wrangler. Root causeThe commitHash variable, | High 7.7 | 20 Jan 2026 |
| CVE-2025-13353 | In gokey versions <0.2.0, a flaw in the seed decryption logic resulted in passwords incorrectly being derived solely from the initial vector and the AES-GCM authentication tag of the key seed. This issue has been fixed in gokey version 0.2.0. This is a breaking change. The fix has invalidated any passwords/secrets that were derived from the seed file (using the -s option). Even if the input seed f | High 7.1 | 2 Dec 2025 |
| CVE-2025-7054 | Cloudflare quiche was discovered to be vulnerable to an infinite loop when sending packets containing RETIRE_CONNECTION_ID frames. QUIC connections possess a set of connection identifiers (IDs); see Section 5.1 of RFC 9000 https://datatracker.ietf.org/doc/html/rfc9000#section-5.1 . Once the QUIC handshake completes, a local endpoint is responsible for issuing and retiring Connection IDs that are u | High 8.7 | 7 Aug 2025 |
| CVE-2025-4821 | Impact Cloudflare quiche was discovered to be vulnerable to incorrect congestion window growth, which could cause it to send data at a rate faster than the path might actually support. An unauthenticated remote attacker can exploit the vulnerability by first completing a handshake and initiating a congestion-controlled data transfer towards itself. Then, it could manipulate the victim's congestion | High 7.5 | 18 Jun 2025 |
| CVE-2025-4820 | Impact Cloudflare quiche was discovered to be vulnerable to incorrect congestion window growth, which could cause it to send data at a rate faster than the path might actually support. An unauthenticated remote attacker can exploit the vulnerability by first completing a handshake and initiating a congestion-controlled data transfer towards itself. Then, it could manipulate the victim's congestion | Medium 5.3 | 18 Jun 2025 |
| CVE-2025-6087 | A Server-Side Request Forgery (SSRF) vulnerability was identified in the @opennextjs/cloudflare package. The vulnerability stems from an unimplemented feature in the Cloudflare adapter for Open Next, which allowed unauthenticated users to proxy arbitrary remote content via the /_next/image endpoint. This issue allowed attackers to load remote resources from arbitrary hosts under the victim site’s | High 7.8 | 16 Jun 2025 |
| CVE-2025-4366 | A request smuggling vulnerability identified within Pingora’s proxying framework, pingora-proxy, allows malicious HTTP requests to be injected via manipulated request bodies on cache HITs, leading to unauthorized request execution and potential cache poisoning. Fixed in: https://github.com/cloudflare/pingora/commit/fda3317ec822678564d641e7cf1c9b77ee3759ff https://github.com/cloudflare/pingora/comm | High 7.4 | 22 May 2025 |
| CVE-2025-4143 | The OAuth implementation in workers-oauth-provider that is part of MCP framework https://github.com/cloudflare/workers-mcp , did not correctly validate that redirect_uri was on the allowed list of redirect URIs for the given client registration. Fixed in: https://github.com/cloudflare/workers-oauth-provider/pull/26 https://github.com/cloudflare/workers-oauth-provider/pull/26 Impact: Under certain | Medium 6.0 | 1 May 2025 |
| CVE-2025-4144 | PKCE was implemented in the OAuth implementation in workers-oauth-provider that is part of MCP framework https://github.com/cloudflare/workers-mcp . However, it was found that an attacker could cause the check to be skipped. Fixed in: https://github.com/cloudflare/workers-oauth-provider/pull/27 https://github.com/cloudflare/workers-oauth-provider/pull/27 Impact: PKCE is a defense-in-depth mechanis | Medium 5.3 | 1 May 2025 |
| CVE-2021-3978 | When copying files with rsync, octorpki uses the "-a" flag 0, which forces rsync to copy binaries with the suid bit set as root. Since the provided service definition defaults to root ( https://github.com/cloudflare/cfrpki/blob/master/package/octorpki.service ) this could allow for a vector, when combined with another vulnerability that causes octorpki to process a malicious TAL file, for a local | High 7.5 | 29 Jan 2025 |
| CVE-2025-0651 | Improper Privilege Management vulnerability in Cloudflare WARP on Windows allows File Manipulation. User with a low system privileges can create a set of symlinks inside the C:\ProgramData\Cloudflare\warp-diag-partials folder. After triggering the 'Reset all settings" option the WARP service will delete the files that the symlink was pointing to. Given the WARP service operates with System privile | Medium 6.1 | 22 Jan 2025 |
| CVE-2024-1765 | Cloudflare Quiche (through version 0.19.1/0.20.0) was affected by an unlimited resource allocation vulnerability causing rapid increase of memory usage of the system running quiche server or client. A remote attacker could take advantage of this vulnerability by repeatedly sending an unlimited number of 1-RTT CRYPTO frames after previously completing the QUIC handshake. Exploitation was possible f | Medium 5.9 | 12 Mar 2024 |
| CVE-2024-1410 | Cloudflare quiche was discovered to be vulnerable to unbounded storage of information related to connection ID retirement, which could lead to excessive resource consumption. Each QUIC connection possesses a set of connection Identifiers (IDs); see RFC 9000 Section 5.1 https://datatracker.ietf.org/doc/html/rfc9000#section-5.1 . Endpoints declare the number of active connection IDs they are willing | Low 3.7 | 12 Mar 2024 |
| CVE-2024-0212 | The Cloudflare Wordpress plugin was found to be vulnerable to improper authentication. The vulnerability enables attackers with a lower privileged account to access data from the Cloudflare API. | High 8.1 | 29 Jan 2024 |
| CVE-2023-6992 | Cloudflare version of zlib library was found to be vulnerable to memory corruption issues affecting the deflation algorithm implementation (deflate.c). The issues resulted from improper input validation and heap-based buffer overflow. A local attacker could exploit the problem during compression using a crafted malicious file potentially leading to denial of service of the software. Patches: The i | Medium 4.0 | 4 Jan 2024 |
| CVE-2023-7078 | Sending specially crafted HTTP requests to Miniflare's server could result in arbitrary HTTP and WebSocket requests being sent from the server. If Miniflare was configured to listen on external network interfaces (as was the default in wrangler until 3.19.0), an attacker on the local network could access other local servers. | High 7.5 | 29 Dec 2023 |
| CVE-2023-7079 | Sending specially crafted HTTP requests and inspector messages to Wrangler's dev server could result in any file on the user's computer being accessible over the local network. An attacker that could trick any user on the local network into opening a malicious website could also read any file. | Medium 6.4 | 29 Dec 2023 |
| CVE-2023-7080 | The V8 inspector intentionally allows arbitrary code execution within the Workers sandbox for debugging. wrangler dev would previously start an inspector server listening on all network interfaces. This would allow an attacker on the local network to connect to the inspector and run arbitrary code. Additionally, the inspector server did not validate Origin/Host headers, granting an attacker that c | High 8.5 | 29 Dec 2023 |
| CVE-2023-6193 | quiche v. 0.15.0 through 0.19.0 was discovered to be vulnerable to unbounded queuing of path validation messages, which could lead to excessive resource consumption. QUIC path validation (RFC 9000 Section 8.2) requires that the recipient of a PATH_CHALLENGE frame responds by sending a PATH_RESPONSE. An unauthenticated remote attacker can exploit the vulnerability by sending PATH_CHALLENGE frames a | Medium 5.3 | 12 Dec 2023 |
| CVE-2023-6180 | The tokio-boring library in version 4.0.0 is affected by a memory leak issue that can lead to excessive resource consumption and potential DoS by resource exhaustion. The set_ex_data function used by the library did not deallocate memory used by pre-existing data in memory each time after completing a TLS connection causing the program to consume more resources with each new connection. | Medium 5.3 | 5 Dec 2023 |
| CVE-2023-3747 | Zero Trust Administrators have the ability to disallow end users from disabling WARP on their devices. Override codes can also be created by the Administrators to allow a device to temporarily be disconnected from WARP, however, due to lack of server side validation, an attacker with local access to the device, could extend the maximum allowed disconnected time of WARP client granted by an overrid | Medium 5.5 | 7 Sep 2023 |
| CVE-2023-0654 | Due to a misconfiguration, the WARP Mobile Client (< 6.29) for Android was susceptible to a tapjacking attack. In the event that an attacker built a malicious application and managed to install it on a victim's device, the attacker would be able to trick the user into believing that the app shown on the screen was the WARP client when in reality it was the attacker's app. | Low 3.9 | 29 Aug 2023 |
| CVE-2023-0238 | Due to lack of a security policy, the WARP Mobile Client (<=6.29) for Android was susceptible to this vulnerability which allowed a malicious app installed on a victim's device to exploit a peculiarity in an Android function, wherein under certain conditions, the malicious app could dictate the task behaviour of the WARP app. | Low 3.9 | 29 Aug 2023 |
| CVE-2023-4241 | lol-html can cause panics on certain HTML inputs. Anyone processing arbitrary 3rd party HTML with the library is affected. | High 7.5 | 16 Aug 2023 |
| CVE-2023-3766 | A vulnerability was discovered in the odoh-rs rust crate that stems from faulty logic during the parsing of encrypted queries. This issue specifically occurs when processing encrypted query data received from remote clients and enables an attacker with knowledge of this vulnerability to craft and send specially designed encrypted queries to targeted ODOH servers running with odoh-rs. Upon successf | Medium 5.9 | 3 Aug 2023 |
| CVE-2023-3348 | The Wrangler command line tool (<[email protected] or <[email protected]) was affected by a directory traversal vulnerability when running a local development server for Pages (wrangler pages dev command). This vulnerability enabled an attacker in the same network as the victim to connect to the local development server and access the victim's files present outside of the directory for the developmen | Medium 5.7 | 3 Aug 2023 |
| CVE-2023-2754 | The Cloudflare WARP client for Windows assigns loopback IPv4 addresses for the DNS Servers, since WARP acts as local DNS server that performs DNS queries in a secure manner, however, if a user is connected to WARP over an IPv6-capable network, te WARP client did not assign loopback IPv6 addresses but Unique Local Addresses, which under certain conditions could point towards unknown devices in the | High 7.4 | 3 Aug 2023 |
| CVE-2023-1862 | Cloudflare WARP client for Windows (up to v2023.3.381.0) allowed a malicious actor to remotely access the warp-svc.exe binary due to an insufficient access control policy on an IPC Named Pipe. This would have enabled an attacker to trigger WARP connect and disconnect commands, as well as obtaining network diagnostics and application configuration from the target's device. It is important to note t | High 7.3 | 20 Jun 2023 |
| CVE-2023-3036 | An unchecked read in NTP server in github.com/cloudflare/cfnts prior to commit 783490b https://github.com/cloudflare/cfnts/commit/783490b913f05e508a492cd7b02e3c4ec2297b71 enabled a remote attacker to trigger a panic by sending an NTSAuthenticator packet with extension length longer than the packet contents. | High 8.6 | 14 Jun 2023 |
| CVE-2023-3040 | A debug function in the lua-resty-json package, up to commit id 3ef9492bd3a44d9e51301d6adc3cd1789c8f534a (merged in PR #14) contained an out of bounds access bug that could have allowed an attacker to launch a DoS if the function was used to parse untrusted input data. It is important to note that because this debug function was only used in tests and demos, it was not exploitable in a normal envi | Low 3.7 | 14 Jun 2023 |
| CVE-2023-2512 | Prior to version v1.20230419.0, the FormData API implementation was subject to an integer overflow. If a FormData instance contained more than 2^31 elements, the forEach() method could end up reading from the wrong location in memory while iterating over elements. This would most likely lead to a segmentation fault, but could theoretically allow arbitrary undefined behavior. In order for the bug t | Medium 6.5 | 12 May 2023 |
| CVE-2023-1732 | When sampling randomness for a shared secret, the implementation of Kyber and FrodoKEM, did not check whether crypto/rand.Read() returns an error. In rare deployment cases (error thrown by the Read() function), this could lead to a predictable shared secret. The tkn20 and blindrsa components did not check whether enough randomness was returned from the user provided randomness source. Typically th | Medium 5.3 | 10 May 2023 |
| CVE-2023-0652 | Due to a hardlink created in the ProgramData folder during the repair process of the software, the installer (MSI) of WARP Client for Windows (<= 2022.12.582.0) allowed a malicious attacker to forge the destination of the hardlink and escalate privileges, overwriting SYSTEM protected files. As Cloudflare WARP client for Windows (up to version 2022.5.309.0) allowed creation of mount points from its | High 7.0 | 6 Apr 2023 |
| CVE-2023-1412 | An unprivileged (non-admin) user can exploit an Improper Access Control vulnerability in the Cloudflare WARP Client for Windows (<= 2022.12.582.0) to perform privileged operations with SYSTEM context by working with a combination of opportunistic locks (oplock) and symbolic links (which can both be created by an unprivileged user). After installing the Cloudflare WARP Client (admin privileges requ | High 7.0 | 5 Apr 2023 |
| CVE-2023-1314 | A vulnerability has been discovered in cloudflared's installer (<= 2023.3.0) for Windows 32-bits devices that allows a local attacker with no administrative permissions to escalate their privileges on the affected device. This vulnerability exists because the MSI installer used by cloudflared relied on a world-writable directory. An attacker with local access to the device (without Administrator r | High 7.5 | 21 Mar 2023 |
| CVE-2022-4428 | support_uri parameter in the WARP client local settings file (mdm.xml) lacked proper validation which allowed for privilege escalation and launching an arbitrary executable on the local machine upon clicking on the "Send feedback" option. An attacker with access to the local file system could use a crafted XML config file pointing to a malicious file or set a local path to the executable using Clo | High 8.9 | 11 Jan 2023 |
| CVE-2022-4457 | Due to a misconfiguration in the manifest file of the WARP client for Android, it was possible to a perform a task hijacking attack. An attacker could create a malicious mobile application which could hijack legitimate app and steal potentially sensitive information when installed on the victim's device. | Medium 5.5 | 11 Jan 2023 |
| CVE-2014-125026 | LZ4 bindings use a deprecated C API that is vulnerable to memory corruption, which could lead to arbitrary code execution if called with untrusted user input. | Critical 9.8 | 27 Dec 2022 |
| CVE-2022-3337 | It was possible for a user to delete a VPN profile from WARP mobile client on iOS platform despite the Lock WARP switch https://developers.cloudflare.com/cloudflare-one/connections/connect-devices/warp/warp-settings/#lock-warp-switch feature being enabled on Zero Trust Platform. This led to bypassing policies and restrictions enforced for enrolled devices by the Zero Trust platform. | Medium 6.7 | 28 Oct 2022 |
| CVE-2022-3512 | Using warp-cli command "add-trusted-ssid", a user was able to disconnect WARP client and bypass the "Lock WARP switch" feature resulting in Zero Trust policies not being enforced on an affected endpoint. | Medium 6.7 | 28 Oct 2022 |
| CVE-2022-3321 | It was possible to bypass Lock WARP switch feature https://developers.cloudflare.com/cloudflare-one/connections/connect-devices/warp/warp-settings/#lock-warp-switch on the WARP iOS mobile client by enabling both "Disable for cellular networks" and "Disable for Wi-Fi networks" switches at once in the application settings. Such configuration caused the WARP client to disconnect and allowed the user | Medium 6.7 | 28 Oct 2022 |
| CVE-2022-3322 | Lock Warp switch is a feature of Zero Trust platform which, when enabled, prevents users of enrolled devices from disabling WARP client. Due to insufficient policy verification by WARP iOS client, this feature could be bypassed by using the "Disable WARP" quick action. | Medium 6.7 | 28 Oct 2022 |
| CVE-2022-3320 | It was possible to bypass policies configured for Zero Trust Secure Web Gateway by using warp-cli 'set-custom-endpoint' subcommand. Using this command with an unreachable endpoint caused the WARP Client to disconnect and allowed bypassing administrative restrictions on a Zero Trust enrolled endpoint. | Medium 6.7 | 28 Oct 2022 |
| CVE-2022-3616 | Attackers can create long chains of CAs that would lead to OctoRPKI exceeding its max iterations parameter. In consequence it would cause the program to crash, preventing it from finishing the validation and leading to a denial of service. Credits to Donika Mirdita and Haya Shulman - Fraunhofer SIT, ATHENE, who discovered and reported this vulnerability. | Medium 5.4 | 28 Oct 2022 |
| CVE-2022-2529 | sflow decode package does not employ sufficient packet sanitisation which can lead to a denial of service attack. Attackers can craft malformed packets causing the process to consume large amounts of memory resulting in a denial of service. | High 7.5 | 30 Sep 2022 |
| CVE-2022-2225 | By using warp-cli subcommands (disable-ethernet, disable-wifi), it was possible for a user without admin privileges to bypass configured Zero Trust security policies (e.g. Secure Web Gateway policies) and features such as 'Lock WARP switch'. | High 8.1 | 26 Jul 2022 |
| CVE-2022-2145 | Cloudflare WARP client for Windows (up to v. 2022.5.309.0) allowed creation of mount points from its ProgramData folder. During installation of the WARP client, it was possible to escalate privileges and overwrite SYSTEM protected files. | Medium 5.8 | 28 Jun 2022 |
| CVE-2022-2147 | Cloudflare Warp for Windows from version 2022.2.95.0 contained an unquoted service path which enables arbitrary code execution leading to privilege escalation. The fix was released in version 2022.3.186.0. | Medium 6.5 | 23 Jun 2022 |
| CVE-2021-3907 | OctoRPKI does not escape a URI with a filename containing "..", this allows a repository to create a file, (ex. rsync://example.org/repo/../../etc/cron.daily/evil.roa), which would then be written to disk outside the base cache folder. This could allow for remote code execution on the host machine OctoRPKI is running on. | High 7.4 | 11 Nov 2021 |
| CVE-2021-3908 | OctoRPKI does not limit the depth of a certificate chain, allowing for a CA to create children in an ad-hoc fashion, thereby making tree traversal never end. | Medium 5.9 | 11 Nov 2021 |
| CVE-2021-3909 | OctoRPKI does not limit the length of a connection, allowing for a slowloris DOS attack to take place which makes OctoRPKI wait forever. Specifically, the repository that OctoRPKI sends HTTP requests to will keep the connection open for a day before a response is returned, but does keep drip feeding new bytes to keep the connection alive. | Medium 4.4 | 11 Nov 2021 |
| CVE-2021-3910 | OctoRPKI crashes when encountering a repository that returns an invalid ROA (just an encoded NUL (\0) character). | Medium 4.4 | 11 Nov 2021 |
| CVE-2021-3911 | If the ROA that a repository returns contains too many bits for the IP address then OctoRPKI will crash. | Medium 4.2 | 11 Nov 2021 |
| CVE-2021-3912 | OctoRPKI tries to load the entire contents of a repository in memory, and in the case of a GZIP bomb, unzip it in memory, making it possible to create a repository that makes OctoRPKI run out of memory (and thus crash). | Medium 4.2 | 11 Nov 2021 |
| CVE-2021-3761 | Any CA issuer in the RPKI can trick OctoRPKI prior to 1.3.0 into emitting an invalid VRP "MaxLength" value, causing RTR sessions to terminate. An attacker can use this to disable RPKI Origin Validation in a victim network (for example AS 13335 - Cloudflare) prior to launching a BGP hijack which during normal operations would be rejected as "RPKI invalid". Additionally, in certain deployments RTR s | High 7.5 | 9 Sep 2021 |
Original policy
Copied from sorona.xyz on 12 Nov 2026. The summaries on this page are made from it; the live page may have changed since.
Skip to content What this pays sorona.xyz pays 1.25× the program's reward, up to +$1,000 per finding, plus a complexity bonus of up to $1,000 for especially complex or well-chained bugs. Finding │ Program pays │ Through sorona.xyz Critical (workerd's own source-code asset, default row) │ $10,000 │ $11,000 High │ $3,000 │ $3,750 Medium │ $750 │ $938 Low │ $500 │ $625 Billing and plan-enforcement issues (capped by policy) │ $250 │ $313 Submit a vulnerability CVE history 0 Last 12 months 2 Lifetime Access control 30 % Info disclosure 25 % Code execution 25 % Denial of service 15 % Latest advisories Loading the feed Overview workerd is the actual runtime binary behind Cloudflare Workers, which serves a double-digit share of all HTTP requests on the public internet through Cloudflare's edge, and it is equally the engine under Pages Functions, Durable Objects, Workers AI gateway paths and every developer's wrangler dev . Third parties embed it independently of Cloudflare when they want a V8-isolate-per-request server, and it is the reference implementation the WinterCG/WinterTC server-side JavaScript standards track is written against. Because the isolate model is multi-tenant, a single escape is a cross-customer event, which is why Cloudflare put workerd on its own source-code scope at the default reward row, $500 at Low up to $10,000 for a Critical, rather than the reduced open-source band. sorona.xyz adds 1.25x on top of whatever Cloudflare awards, up to $1,000 more, plus a complexity bonus of up to $1,000 for a well-chained finding. Response times 10 business days First response 10 business days after first response Triage 10 business days after triage Bounty In scope workerd is one of the few Cloudflare repositories the program names on its own. The live scope table carries github.com/cloudflare/workerd as a SOURCE_CODE asset, bounty-eligible at a maximum severity of critical, with no binding to the reduced open-source row, so the default ladder up to $10,000 applies. The declaration is the repository itself rather than a hosted wrapper around it, and it covers the runtime as a whole The workerd server binary: process startup, socket and service wiring, and the Cap'n Proto configuration and control plane an embedder feeds untrusted config into The binding layer that glues V8 to the C++ runtime, where a type or lifetime mistake crosses the JavaScript boundary The whole Workers runtime API surface: fetch , Request / Response , Headers , streams, WebSocket , the Cache API, WebCrypto, URL and text encoding, timers and everything else a Worker can call V8 isolate lifecycle management, isolate boundaries, and any state that survives a request or crosses between tenants sharing a machine Durable Objects in full: runtime semantics, the storage API, alarms, and the isolation between clients and sessions talking to the same object The bindings workerd implements for a Worker to reach other services, including service bindings, queues and the KV, R2 and D1 interfaces the runtime exposes WebAssembly hosting inside a Worker, including the boundary between guest module and host runtime The Node.js compatibility shims exposed under nodejs_compat , where a shim can hand back more capability than the real API would The inspector and devtools endpoint workerd opens, and the local development paths wrangler dev drives through it Out of scope V8 itself, which is Chrome VRP territory and already sits on the sorona.xyz paid rail as its own program Local-only wrangler dev reproductions with no demonstrated production reachability The repositories Cloudflare marks explicitly not eligible for bounty: cloudflare/vinext-private , moq-rs , privacy-gateway-server-go , saffron and recapn Billing and plan-enforcement issues, which the policy caps at Low severity and a $250 maximum payout Findings that require a maintainer to approve a GitHub Actions workflow before they trigger Reports with no working proof of concept, which are not bounty-eligible at any severity Most-valued findings An isolate escape or cross-request state leak that lets one tenant's Worker read or influence another's, the finding that carries the $10,000 Critical row Durable Objects isolation failures where one client reaches another client's object state or session Code execution reachable from Worker-controlled input into the host runtime, including through the WebAssembly host boundary Access-control failures in the Cap'n Proto control plane that let untrusted configuration change what a running workerd will serve nodejs_compat shims that expose filesystem, process or network capability the Workers sandbox is supposed to withhold Any of the above shown reachable in production rather than only in a local build, which is the difference between a bounty and an Informational close Program rules Submit through sorona.xyz first. We validate, coordinate with the program's security team, and add the payout uplift on top of whatever it awards. A working proof of concept is required: no PoC, no bounty, and source-code-only analysis is Informational regardless of the severity you assign it A local workerd build or wrangler dev reproduction is graded Informational unless you show the same code path is reachable in production Cloudflare runs its own AI-assisted discovery and closes matching external reports as duplicates, so file early rather than polishing Billing and plan-enforcement issues are capped at Low severity and $250 maximum payout however you frame them The generic github.com/cloudflare catch-all scope pays a lower $250 to $3,500 band; the figures below apply because workerd is separately named Program stats 376 Reports resolved $568k Bounties paid 77% Response efficiency Disclosure Coordinated with Cloudflare on a bounded clock: nothing may be published for 90 days from submission without prior written approval. The program is unusually well documented, with published response, triage and bounty targets, so sorona.xyz files the report, tracks it against those targets, and holds publication until the window closes. Safe harbor Sorona will not pursue legal action against you for good-faith research submitted to us in line with this policy. We cannot speak for the upstream project, its vendor, or any third party, and we do not grant authorization on their behalf. You are responsible for understanding and following the laws that apply to you, wherever you are, and for staying within any scope the project has defined. Submit a vulnerability