New MCP server — Bring vulnerability intelligence into your AI
Back to feed

CVE-2026-47406

NVDCIRCLENISA

EUVD-2026-46315 · Published May 29, 2026

praisonai-platform: IDOR in dependency endpoints allows cross-workspace issue linking, reading, and deletion due to missing ownership checks

Summary

Type: Insecure Direct Object Reference. The dependency endpoints (POST/GET /workspaces/{workspace_id}/issues/{issue_id}/dependencies and DELETE .../dependencies/{dep_id}) gate access on require_workspace_member(workspace_id) only, then dispatch to DependencyService calls that take URL/body-supplied issue and dependency IDs without verifying any of them belong to the membership-checked workspace. Most damaging: create_dependency accepts body.depends_on_issue_id from the request body — that ID is checked against nothing — letting an attacker create a "blocks" or "related" link between any two issues anywhere in the database.
File: src/praisonai-platform/praisonai_platform/api/routes/dependencies.py, lines 22-58; services/dependency_service.py, lines 26-65.
Root cause: the same Depends(require_workspace_member) default-min-role pattern as the companion IDORs, plus a service layer (DependencyService) where every method takes raw IDs and queries them directly. create(issue_id, depends_on_issue_id, ...) writes a row with no workspace verification on either ID. list_for_issue(issue_id) returns dependencies in either direction. delete(dep_id) is a primary-key delete with no workspace predicate.

Affected Code

File 1: src/praisonai-platform/praisonai_platform/api/routes/dependencies.py, lines 22-58.

@router.post("/", response_model=DependencyResponse, status_code=status.HTTP_201_CREATED)
async def create_dependency(
    workspace_id: str,
    issue_id: str,
    body: DependencyCreate,
    user: AuthIdentity = Depends(require_workspace_member),
    session: AsyncSession = Depends(get_db),
):
    svc = DependencyService(session)
    dep = await svc.create(issue_id, body.depends_on_issue_id, body.type)  # <-- BUG: neither id is workspace-checked
    return DependencyResponse.model_validate(dep)


@router.get("/", response_model=List[DependencyResponse])
async def list_dependencies(
    workspace_id: str,
    issue_id: str,
    user: AuthIdentity = Depends(require_workspace_member),
    session: AsyncSession = Depends(get_db),
):
    svc = DependencyService(session)
    deps = await svc.list_for_issue(issue_id)                              # <-- BUG: returns dependencies for any issue
    return [DependencyResponse.model_validate(d) for d in deps]


@router.delete("/{dep_id}", status_code=status.HTTP_204_NO_CONTENT)
async def delete_dependency(
    workspace_id: str,
    issue_id: str,
    dep_id: str,
    user: AuthIdentity = Depends(require_workspace_member),
    session: AsyncSession = Depends(get_db),
):
    svc = DependencyService(session)
    deleted = await svc.delete(dep_id)                                     # <-- BUG: deletes any dependency by id
    if not deleted:
        raise HTTPException(status_code=404, detail="Dependency not found")

File 2: src/praisonai-platform/praisonai_platform/services/dependency_service.py, lines 26-65.

async def create(self, issue_id: str, depends_on_issue_id: str, dep_type: str = "blocks") -> IssueDependency:
    if dep_type not in VALID_TYPES:
        raise ValueError(...)
    dep = IssueDependency(
        issue_id=issue_id,                                              # <-- accepts any
        depends_on_issue_id=depends_on_issue_id,                        # <-- accepts any (from request body)
        type=dep_type,
    )
    self._session.add(dep); await self._session.flush(); return dep

async def list_for_issue(self, issue_id: str) -> list[IssueDependency]:
    stmt = select(IssueDependency).where(
        (IssueDependency.issue_id == issue_id) | (IssueDependency.depends_on_issue_id == issue_id)
    )
    return list((await self._session.execute(stmt)).scalars().all())

async def delete(self, dep_id: str) -> bool:
    dep = await self.get(dep_id)                                        # session.get(IssueDependency, dep_id) — no workspace check
    ...

Why it's wrong: the request-body depends_on_issue_id is the worst part: an attacker can link any two issues across any two workspaces, polluting both workspaces' dependency graphs with attacker-chosen relationships ("blocks", "blocked_by", "related"). The triagers in the foreign workspace see their issue suddenly blocked by an unrelated foreign issue, breaking sprint planning and creating false correlation. The delete(dep_id) path lets an attacker remove legitimate cross-issue links between any two foreign workspaces, also disrupting their planning. The list_for_issue path leaks the dependency graph for any issue in the deployment.

Exploit Chain

  1. Attacker is a member of workspace W_attacker and harvests two foreign-workspace issue UUIDs I1 (in W_target1) and I2 (in W_target2). They leak via the activity feed, comment threads, error messages, exported dumps, the agent prompt history, or any other channel that ever serialises an issue ID. State: attacker holds two foreign issue UUIDs.
  2. Attacker sends POST /workspaces/W_attacker/issues/I1/dependencies with Authorization: Bearer <attacker_jwt> and body {"depends_on_issue_id": "I2", "type": "blocks"}. State: control flow enters create_dependency with issue_id=I1 (foreign), depends_on_issue_id=I2 (foreign).
  3. require_workspace_member(W_attacker, attacker) passes (attacker is a member of W_attacker). DependencyService.create(I1, I2, "blocks") writes a new row IssueDependency(issue_id=I1, depends_on_issue_id=I2, type="blocks"). State: there is now a cross-workspace dependency between two foreign issues, written by the attacker.
  4. The triage UIs of W_target1 and W_target2 now show that the foreign issue is blocked by an unrelated issue in another workspace. Workflow rules that key off "cannot close while blocked" will refuse to let the legitimate triagers close I1. State: foreign workflow disrupted.
  5. Attacker repeats with GET /workspaces/W_attacker/issues/I1/dependencies to read the dependency graph for any foreign issue (information disclosure, project relationship mapping), or with DELETE .../{dep_id} (after enumerating dep_ids via the list call) to strip legitimate dependencies between foreign issues, breaking blocked-by chains.
  6. Final state: with one workspace-member token, the attacker reads, writes, and deletes dependencies on every issue in the multi-tenant deployment, polluting the dependency graphs of foreign workspaces.

Security Impact

Severity: sec-high. CVSS 7.6: network attack, low complexity, low privileges, no user interaction, scope unchanged, high confidentiality (cross-workspace dependency graph disclosure), high integrity (cross-workspace dependency injection and deletion), no availability claim (workflow disruption is integrity, not availability).
Attacker capability: read any issue's dependency graph; create arbitrary "blocks" / "blocked_by" / "related" links between any two issues across any two workspaces; delete any dependency by id. The most surprising primitive is the cross-workspace LINKING — the only one of the IDORs in this codebase where a single attacker request can affect TWO foreign workspaces at once.
Preconditions: praisonai-platform is deployed multi-tenant; attacker has any membership token; foreign issue UUIDs are reachable.
Differential: source-inspection-verified end-to-end. The asymmetry between this service (no workspace predicate anywhere) and MemberService.get(workspace_id, user_id) (correctly composite-keyed) confirms the gap. With the suggested fix below, the route would resolve both the URL issue_id and the body depends_on_issue_id against IssueService.get(workspace_id, ...) before allowing the dependency to be written.

Suggested Fix

Resolve every issue id (URL and body) against workspace_id at the route layer before dispatching. The route helper from the issue-IDOR companion advisory can be reused.

--- a/src/praisonai-platform/praisonai_platform/api/routes/dependencies.py
+++ b/src/praisonai-platform/praisonai_platform/api/routes/dependencies.py
@@ -22,11 +22,16 @@
 @router.post("/", response_model=DependencyResponse, status_code=status.HTTP_201_CREATED)
 async def create_dependency(
     workspace_id: str,
     issue_id: str,
     body: DependencyCreate,
     user: AuthIdentity = Depends(require_workspace_member),
     session: AsyncSession = Depends(get_db),
 ):
+    issue_svc = IssueService(session)
+    if await issue_svc.get(workspace_id, issue_id) is None:
+        raise HTTPException(status_code=404, detail="Issue not found")
+    if await issue_svc.get(workspace_id, body.depends_on_issue_id) is None:
+        raise HTTPException(status_code=404, detail="depends_on_issue_id not found in this workspace")
     svc = DependencyService(session)
     dep = await svc.create(issue_id, body.depends_on_issue_id, body.type)
     return DependencyResponse.model_validate(dep)

Apply the same issue_svc.get(workspace_id, issue_id) precondition to list_dependencies and delete_dependency (verifying both the issue and the dependency belong to workspace_id).

Technologies

Praisonai-Platform Praisonai-Platform

Weaknesses

CWE-639

CVSS scores

  • v3.18.1CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N

MITRE ATT&CK

Inferred

Where this vulnerability fits in an attack

Initial Access

References

Potentially impacted assets

See if this affects your attack surface

Latest trending attack

Criticalyesterday

Google Brotli decoder - Remote Code Execution via compressed requests (Exploitation Check)

This control is the second step of the Patrowl response to CVE-2026-111497 (Node2Shell). It complements the first control of 22 October 2026, which focused on the Node.js runtime, and does not replace it. Why a second control CVE-2026-111497 is an out-of-bounds write in the Google Brotli decoder, the library that decompresses Brotli ("br") data. It was first published as a Node.js vulnerability, because Node.js bundles this decoder. Later publications showed that the flaw is in the library itself. Every product that embeds an affected version of the decoder is concerned, whatever its language or vendor. A public proof of concept demonstrates code execution. Any server that decompresses request bodies sent with "Content-Encoding: br" exposes the decoder to unauthenticated attackers. Known cases include Express 5 and NestJS 11 on Node.js, Koa and Fastify setups, Bun, Envoy and Istio with request decompression enabled, and any application or reverse proxy that decodes Brotli request bodies with an affected library. The flaw also works in the other direction. A malicious server can compromise an unpatched client that decodes a Brotli response, for example Node.js fetch(), Python clients (requests, httpx, aiohttp, Scrapy) with the Brotli or brotlicffi package, or libcurl builds with Brotli support. What this control checks Patrowl sends small, valid Brotli-compressed requests to each asset and observes whether the server decodes them. The check does not depend on the technology: it detects the behaviour, whatever the stack behind it. These requests are what any legitimate client could send. They never reach the vulnerable code path and cannot affect the availability of the service. How to read the result Potentially impacted means the asset decodes Brotli request bodies: the attack surface is exposed. It does not confirm that the asset is vulnerable. The version of the Brotli decoder in use is not visible from the outside. This check covers applications that parse request bodies on their root path. A Not impacted result does not rule out exposure on other routes, nor on the client side. Checking the versions on the servers remains the only exhaustive verification. Affected and fixed versions Google Brotli 1.0.3 to 1.2.0 are affected, fixed in 1.0.10, 1.1.1 and 1.2.1. Fixed releases of the main products embedding it: Node.js 22.23.4, 24.22.1 and 26.12.1 (end-of-life lines 20.x and older and 25.x receive no fix), Bun 1.4.4, Envoy 1.37.8, 1.38.6, 1.39.3 and 1.40.0, Istio 1.29.9, 1.30.6 and 1.31.2, Python Brotli 1.2.1 and brotlicffi 1.2.1.0. Products that link against the system library depend on the Brotli package of the Linux distribution. Recommendations Update every component that embeds the Brotli decoder (only real fix). Start with the assets flagged by this control. Restart the processes after the update. Rebuild containers, single-executable applications and appliances that bundle an affected runtime or library. Reduce the exposure until the update is deployed. Reject requests with "Content-Encoding: br" at the reverse proxy, load balancer or WAF, case-insensitively and including lists such as "gzip, br". On Envoy and Istio, disable request decompression for Brotli. Review outbound calls. List the features that fetch content from servers you do not control, and patch the services that host them first.

Take 15 minutes to discover our platform with our experts

Patrowl Intel platform screenshot