Summary
An information disclosure flaw in AgentChat v2.3.0 exposes sensitive user details, including SHA256 password hashes, through the /api/v1/user/info endpoint.
Unauthenticated attackers can enumerate user IDs to harvest this data, increasing the risk of account compromise and credential attacks.
Affected context
The vulnerability appears in AgentChat v2.3.0 within the /api/v1/user/info API endpoint.
It manifests when the service returns detailed user records, including password hash values, without enforcing authentication or authorization checks.
Environments that expose this endpoint to untrusted networks and use predictable or enumerable user ID values are particularly susceptible.
Exploit preconditions
- Attack surface: Publicly accessible HTTP(S) API endpoint at
/api/v1/user/info - Required access: No prior authentication; attacker can be an anonymous remote client
- Trigger point: Issuing repeated requests with different user ID values (e.g., sequential or guessable identifiers)
- Why exploitation is possible: The endpoint discloses sensitive fields such as SHA256 password hashes and related user information without validating caller identity or enforcing per-user access controls
Why this happens
The core issue is an overly permissive data exposure pattern in a user information endpoint.
Instead of limiting responses to non-sensitive, minimal profile data scoped to the authenticated subject, the implementation returns sensitive attributes, including password hashes, to any caller.
This behavior aligns with CWE-200: Exposure of Sensitive Information to an Unauthorized Actor, where a system reveals data that should remain confidential due to missing or ineffective access control and data minimization.
Potential impact
- Attackers can obtain SHA256 password hashes for multiple user accounts.
- Exposed hashes may be subjected to offline cracking attempts.
- Compromised credentials could enable account takeover within AgentChat or on other services where users reuse passwords.
- Aggregated user information can support targeted phishing, social engineering, or further privilege escalation attempts.
General mitigation guidance
Mitigation generally centers on tightening access control and minimizing exposed data in API responses.
Common defensive patterns include:
- Enforcing strict authentication and authorization on all user information endpoints.
- Ensuring responses only include non-sensitive, least-privilege profile fields necessary for the client's function.
- Implementing per-user access checks so callers can only retrieve their own data or data within an explicitly authorized scope.
- Separating internal security fields (such as password hashes, tokens, and secrets) from externally facing DTOs or API response models.
- Applying rate limiting and anomaly detection to reduce mass enumeration risks.
How teams detect issues like this
Effective detection combines code-level and runtime perspectives:
- Static analysis: Flag endpoints that serialize and return password hashes, authentication secrets, or credential-like fields in API responses.
- Code review: Look for controllers or handlers that expose ORM entities directly, lack authorization checks, or use generic "return user object" patterns without filtering fields.
- API gateway and WAF policies: Enforce authentication requirements and monitor for patterns indicative of ID enumeration (e.g., high-rate sequential ID requests).
- Logging and monitoring: Track access to user information endpoints and alert on anomalous volume or distribution of requested user IDs.

