Skip to content

Sourced / Security blog

Security blog

Website, API and AI security. Clear scope, useful evidence, and practical fixes.

204 articles / Page 2 of 9

  1. Web App Security

    Set upload budgets across requests and processing

    Define per-file, request, storage, and processing limits. Test streamed uploads, concurrent reservations, partial cleanup, and legitimate near-limit files.

  2. AI & MCP Permissions

    Budget Agent Tool Operations

    Set per-task limits for tool calls, external spend, data volume, retries, and writes, then test loop stopping.

  3. Tenant Isolation & API Authorization

    Budget API Rate Limits per Tenant

    Separate tenant, user, and global API limits so one customer cannot exhaust another.

  4. Tenant Isolation & API Authorization

    Build a Tenant Authorization Matrix

    Define allowed and denied actions for each role, resource, and tenant relationship.

  5. Tenant Isolation & API Authorization

    Carry Tenant Context Into Async Jobs

    Bind each background job to a trusted tenant and recheck permission before sensitive work.

  6. Dependency security

    Check whether an SBOM covers the released software

    Compare an SBOM with packaged components and a defined acceptance profile. Track missing identities, versions, suppliers, relationships, and coverage gaps.

  7. Tenant Isolation & API Authorization

    Check Tenant Notification Recipients

    Verify that each notification reaches only current, authorized recipients.

  8. Tenant Isolation & API Authorization

    Check Tenant Scope in SQL Joins

    Review joins between tenant-owned tables and test mixed-tenant identifiers.

  9. Web App Security

    Check browser origins before WebSocket upgrades

    Define a WebSocket handshake policy for browser and machine clients. Test exact origins, missing origin handling, authentication, and proxy behavior.

  10. Tenant Isolation & API Authorization

    Choose API Denial Responses

    Choose 401, 403, or concealed 404 responses without leaking private resource details.

  11. AI & MCP Permissions

    Constrain MCP Sampling Requests

    Review legacy MCP sampling requests for user approval, data limits, provider destination, and migration from deprecated sampling.

  12. Web App Security

    Constrain URL fetches in import and preview features

    Define which destinations an application may fetch. Review URL parsing, resolved addresses, redirects, response limits, and the worker's network access.

  13. Tenant Isolation & API Authorization

    Constrain Tenant Admin Overrides

    Define narrow, logged exceptions to normal tenant authorization.

  14. AI & MCP Permissions

    Contain Untrusted AI Tool Output

    Keep retrieved pages, tool errors, and structured results from changing an agent's instructions or tool permissions.

  15. Tenant Isolation & API Authorization

    Control Support Impersonation

    Require a named, narrow, time-limited support grant and an audit trail for every action.

  16. Dependency security

    Document a security backport when an upgrade is blocked

    Evaluate a backport against a normal corrective upgrade. Record the blocker, patch identity, maintenance owner, regression evidence, and exit plan.

  17. Credential & Access Lifecycle

    Define Credential Ownership

    Assign each production credential a named owner, backup owner, rotation method, incident contact, and orphaned-key decision.

  18. Tenant Isolation & API Authorization

    Design Synthetic Tenant Test Fixtures

    Build safe test records that reveal both cross-tenant leaks and false denials.

  19. Tenant Isolation & API Authorization

    Design a Tenant Role Permission Map

    Translate tenant roles into explicit resource and action decisions.

  20. Dependency security

    Detect manifest and lockfile drift before release

    Verify that a clean install respects the reviewed dependency graph. Record package-manager settings, workspace scope, install scripts, and unexpected lockfile changes.

  21. Web App Security

    Review output encoding where user content renders

    Trace user-controlled values into text, attributes, links, and rich content. Build rendering assertions that preserve valid text without allowing execution.

  22. AI & MCP Permissions

    Enforce RAG Retrieval Access Controls

    Apply current user and tenant permissions before RAG chunks enter model context, including snippets and caches.

  23. Dependency security

    Expire Dependency Risk Exceptions

    Set a time limit and recheck trigger for dependency waivers so temporary risk decisions do not become permanent defaults.

  24. Tenant Isolation & API Authorization

    Expire Tenant Signed Downloads

    Set issue, use, expiry, and revocation rules for signed tenant file links, then verify them at storage and edge.

Security blog - Page 2 | Sourced