- Top reward
- $1,500
- Top severity paid
- Critical
- Assets in scope
- 15
- Public CVEs
- 10
- Last change
- No changes yet
At a glance
What to know before you start testing.
- Pays from $1,500
- for critical bugs.
- Report on Sorona
- Submit through the program page.
- Partial safe harbor
- Some legal protection; read the terms.
- Publish after a fix
- Coordinate the timing with the company.
Rewards
What Sorona lists as the payout for each severity.
| Severity | Relative size | Reward |
|---|---|---|
| Critical | $1,500+ | |
| High | $700+ | |
| Medium | $300+ | |
| Low |
|
$100 $100 / 1 month Claude Max |
- 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
15 assets you may test, 6 excluded.
Web 1
- https://valkey.io
Products and areas 14
- AUTH, TLS termination, certificate handling and the unix socket path
- CONFIG SET, DEBUG and every command that writes to disk, loads code or changes server behaviour at runtime
- Every value type and its encodings, including the conversion between compact and full representations for hashes, lists, sets, sorted sets, streams and bitmaps
- Key expiry, eviction and keyspace notifications, plus the pub/sub implementation and its channel permission checks
- RDB and AOF file parsing on load, since a server accepts these from restore, from replication and from a rewrite
- Replication end to end, covering the replica handshake, PSYNC, the RDB transfer and the propagation of writes to replicas
- SECURITY.md asks for private reports at [email protected] and describes a coordinated process with vendor pre-notification. It sets no boundary narrower than the server the project releases
- The ACL subsystem: user definitions, command categories, key patterns, channel patterns, selectors and the evaluation that combines them
- The cluster bus protocol, slot assignment, slot migration and the messages nodes exchange about each other
- The command dispatch table and the argument parsing of every command in it
- The Lua scripting engine and the functions engine, their sandboxes, and every library exposed inside them
- The module API and any modules the project ships, including what a loaded module can reach in the server process
- The RESP2 and RESP3 protocol parsers, the inline command path and the connection handshake in front of both
- The Sentinel implementation and the failover decisions it drives
Out of scope (6)
- A server deliberately bound to a public interface with no password and no ACLs, which is a configuration decision
- Client libraries, which are maintained separately
- ElastiCache, Memorystore and other managed services at the infrastructure level, which are the providers' own assets
- Redis itself and Redis Ltd's builds, a separate codebase with its own security process and its own paid engagement pathway
- Resource consumption from commands an authorised user is permitted to run
- Third-party modules that are not shipped in the Valkey tree
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. This project runs no bounty of its own; we validate, coordinate disclosure with the maintainers, and pay you directly.
-
Report privately to the address on the policy page , the address SECURITY.md publishes. GitHub Security Advisories are used for publication
-
Do not open a public GitHub issue. The process runs responsible disclosure with pre-notification to downstream vendors before the advisory
-
Give the exact command sequence and the server configuration, including the ACL rules in force.
-
Say whether the finding needs the scripting engine, a module, or neither.
-
Do not carry a Redis advisory across without re-checking it.
Not accepted
-
Resource consumption from commands an authorised user is permitted to run
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.
Sorona credits this project with 10 CVEs, 6 of them recent. The individual records appear here after the next sync.
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 This project runs no bounty of its own. sorona.xyz pays you directly on qualified, reproduced findings. Severity │ sorona.xyz pays Critical │ $1,500+ High │ $700+ Medium │ $300+ Low │ $100 / 1 month Claude Max Submit a vulnerability CVE history 6 Last 12 months 10 Lifetime Memory safety 40 % Sandbox escape 25 % Access control 20 % Parsing/validation 15 % Latest advisories Loading the feed Overview Valkey is the Linux Foundation fork of Redis, created after the 2024 licence change and now the default Redis-compatible engine in AWS ElastiCache and Google Memorystore. It runs as cache, session store and job queue for an enormous number of applications, holding session tokens and queued work in memory. GHSA-9rfg-jx7v-52p6 is the reference point: a use-after-free in the Lua scripting engine, reachable from an ordinary client connection and ending in code execution on the server. The project's SECURITY.md asks for reports at a mailing list address and mentions no reward of any kind. sorona.xyz pays you directly on qualified, reproduced findings. In scope SECURITY.md asks for private reports at [email protected] and describes a coordinated process with vendor pre-notification. It sets no boundary narrower than the server the project releases The command dispatch table and the argument parsing of every command in it The RESP2 and RESP3 protocol parsers, the inline command path and the connection handshake in front of both The Lua scripting engine and the functions engine, their sandboxes, and every library exposed inside them The ACL subsystem: user definitions, command categories, key patterns, channel patterns, selectors and the evaluation that combines them AUTH, TLS termination, certificate handling and the unix socket path Replication end to end, covering the replica handshake, PSYNC, the RDB transfer and the propagation of writes to replicas The cluster bus protocol, slot assignment, slot migration and the messages nodes exchange about each other RDB and AOF file parsing on load, since a server accepts these from restore, from replication and from a rewrite The module API and any modules the project ships, including what a loaded module can reach in the server process Key expiry, eviction and keyspace notifications, plus the pub/sub implementation and its channel permission checks CONFIG SET , DEBUG and every command that writes to disk, loads code or changes server behaviour at runtime The Sentinel implementation and the failover decisions it drives Every value type and its encodings, including the conversion between compact and full representations for hashes, lists, sets, sorted sets, streams and bitmaps Out of scope Redis itself and Redis Ltd's builds, a separate codebase with its own security process and its own paid engagement pathway ElastiCache, Memorystore and other managed services at the infrastructure level, which are the providers' own assets Third-party modules that are not shipped in the Valkey tree A server deliberately bound to a public interface with no password and no ACLs, which is a configuration decision Client libraries, which are maintained separately Resource consumption from commands an authorised user is permitted to run Most-valued findings Anything reaching code execution from a client connection, the GHSA-9rfg-jx7v-52p6 shape, above all through the Lua or functions engine. A cache server executing attacker code sits inside the application's trust boundary Sandbox escapes out of the scripting engines into the host process, including through the standard library surface exposed to scripts Memory corruption in command argument parsing or in type encoding conversion, since both are reachable by any client that can authenticate ACL evaluation defects where a restricted user runs a command, touches a key pattern or subscribes to a channel the rules never allowed RDB and AOF parsing on load, because a node accepts these files from a replication master, a restore or an operator's backup Cluster bus and replication handshake paths that trust a peer's claimed identity, turning one node into a read of the whole keyspace Commands that write to disk, load modules or change configuration reaching a user who should not hold administrative commands Pre-authentication protocol parsing, since the listener accepts bytes from anything that can reach the port Program rules Submit through sorona.xyz. This project runs no bounty of its own; we validate, coordinate disclosure with the maintainers, and pay you directly. Report privately to [email protected] , the address SECURITY.md publishes. GitHub Security Advisories are used for publication Do not open a public GitHub issue. The process runs responsible disclosure with pre-notification to downstream vendors before the advisory Nothing is paid at any severity. The policy mentions no bounty, reward or swag, and there is nothing to negotiate Give the exact command sequence and the server configuration, including the ACL rules in force. Most severe findings here are reachable from a normal client connection and the ACL state decides whether they are Say whether the finding needs the scripting engine, a module, or neither. The Lua and function engines are the highest-value surface and also the one most often disabled in hardened deployments Do not carry a Redis advisory across without re-checking it. The codebases have diverged since the 2024 fork and Redis has its own separate process Disclosure Coordinated privately with the Valkey maintainers through the security mailing list, with downstream vendors notified ahead of publication so the major cloud providers can patch their managed offerings first. Advisories are published as GHSAs alongside a fixed release. The embargo therefore runs to that coordinated release rather than to a day count you choose. 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