- Top reward
- $1,500
- Top severity paid
- Critical
- Assets in scope
- 13
- Public CVEs
- 0
- 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
13 assets you may test, 6 excluded.
Web 1
- https://typesense.org
Products and areas 12
- Clustering and the Raft-based high-availability path, covering node join, the peer configuration and inter-node requests that arrive already trusted
- Document ingestion and the import path, covering JSON and JSONL parsing, schema validation, auto-schema handling and the write-ahead log
- Every HTTP endpoint the server exposes, including /multi_search, the single-collection search path, documents, collections, aliases, synonyms, overrides, analytics and the keys API
- Scoped search key handling end to end, covering signature verification, the embedded filter that constrains a tenant, and any way a client can widen or strip it
- Search request parsing: filter expressions, faceting, grouping, sorting, vector and hybrid search parameters, and the typo-tolerance machinery itself
- Snapshot and backup handling, which serialises every collection and every key
- The API key system in full: the bootstrap key, derived keys, the actions and collection patterns attached to them, expiry and the scoped search keys generated from a parent key
- The conversational and embedding features, including any outbound request the server makes to a model provider on a user's behalf
- The official client libraries the project publishes
- The on-disk store and index format, including anything that reads back state a previous version wrote
- The published policy asks for any security issue by email to the project security address and draws no component boundary at all. The whole server and the official clients sit inside it
- The search cache and any other cross-request state, including whether a cached result can be served to a request that was not entitled to it
Out of scope (6)
- An application that ships its admin key to the browser instead of generating a scoped search key, which is an integration mistake
- Public GitHub issues, which are not a private channel for a security report
- Running the server on a public network with no key configured, which is an operator decision
- Third-party clients and integrations outside the project's own organisation
- Typesense Cloud's hosting infrastructure rather than the server code
- Upstream dependencies, unless Typesense's own use of them is what makes the defect reachable
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.
-
Do not open a public GitHub issue for a security bug
-
Name the Typesense version and say which key you held: the bootstrap key, a scoped search key, a key with specific actions and collections, or none at all
-
State whether the endpoint was reachable without authentication.
-
Hold detail until the advisory and the release are out.
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.
No CVEs matched yet. We match NVD records by vendor name, so products sold under another name can be missing.
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 2 Last 12 months Denial of service 35 % Cache isolation 30 % API key scoping 25 % Info disclosure 10 % Latest advisories Loading the feed Overview Typesense is the typo-tolerant search engine teams drop in when they want Algolia's behaviour on their own hardware, and it fronts the search box on a large number of production sites and documentation portals. Scoped API keys are its multi-tenant boundary, so isolation between one tenant's documents and another's is enforced entirely in this code. CVE-2026-47216 is the reference point: an unauthenticated denial of service reachable through /multi_search , found and published by the project's own founder. The security policy is one sentence long and contains no reward language of any kind. sorona.xyz pays you directly on qualified, reproduced findings. In scope The published policy asks for any security issue by email to the project security address and draws no component boundary at all. The whole server and the official clients sit inside it The API key system in full: the bootstrap key, derived keys, the actions and collection patterns attached to them, expiry and the scoped search keys generated from a parent key Scoped search key handling end to end, covering signature verification, the embedded filter that constrains a tenant, and any way a client can widen or strip it Every HTTP endpoint the server exposes, including /multi_search , the single-collection search path, documents, collections, aliases, synonyms, overrides, analytics and the keys API Search request parsing: filter expressions, faceting, grouping, sorting, vector and hybrid search parameters, and the typo-tolerance machinery itself The search cache and any other cross-request state, including whether a cached result can be served to a request that was not entitled to it Document ingestion and the import path, covering JSON and JSONL parsing, schema validation, auto-schema handling and the write-ahead log The on-disk store and index format, including anything that reads back state a previous version wrote Clustering and the Raft-based high-availability path, covering node join, the peer configuration and inter-node requests that arrive already trusted Snapshot and backup handling, which serialises every collection and every key The conversational and embedding features, including any outbound request the server makes to a model provider on a user's behalf The official client libraries the project publishes Out of scope Typesense Cloud's hosting infrastructure rather than the server code An application that ships its admin key to the browser instead of generating a scoped search key, which is an integration mistake Running the server on a public network with no key configured, which is an operator decision Third-party clients and integrations outside the project's own organisation Upstream dependencies, unless Typesense's own use of them is what makes the defect reachable Public GitHub issues, which are not a private channel for a security report Most-valued findings Unauthenticated paths that cost the server real work or crash it. CVE-2026-47216 did exactly that through /multi_search , and the search endpoint is the one thing deployments routinely expose to the open internet Cache isolation failures where a result computed for one scoped key is served to another. CVE-2026-47225 is the reference case, and on a multi-tenant install it returns another customer's documents Scoped search key bypass: stripping, widening or forging the embedded filter that is supposed to pin a request to one tenant Anything that recovers the bootstrap key or another key's value from an endpoint a search-only key can reach Memory safety in the C++ index and store layer, particularly anything reachable from a search parameter or from a document an application ingested Filter and query expression parsing that turns a short request into unbounded work or reaches documents the filter excludes Inter-node requests in a high-availability cluster accepted as pre-authenticated from outside, which collapses the key model entirely 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. Email [email protected] , the address in both the repository policy and the vendor security page GitHub private vulnerability reporting is off on typesense/typesense , so email is the only private route Do not open a public GitHub issue for a security bug The channel demonstrably works: two advisories in 2026 were authored and published by the project founder, so expect a technical reader rather than a triage queue Name the Typesense version and say which key you held: the bootstrap key, a scoped search key, a key with specific actions and collections, or none at all State whether the endpoint was reachable without authentication. The most recent high-severity finding here was an unauthenticated denial of service, so pre-auth reachability is the detail that sets severity There is no bounty. The policy is a single sentence and neither it nor the vendor site mentions reward, compensation or payment Disclosure Coordinated privately by email and published as a GitHub Security Advisory with a CVE alongside a fixed release. The 2026 pair went out on the same day, so expect findings to be batched into one advisory release rather than published individually as they are fixed. Hold detail until the advisory and the release are out. 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