Every CVESearch/Get API key

GHSA-f9ff-5x35-7gfw

Grackle: Fail-open authorization in the MCP tool layer lets scoped agents perform cross-task and cross-session mutations (IDOR)

8.7 High

GHSA-f9ff-5x35-7gfw is a high vulnerability (CVSS 8.7) in @grackle-ai/mcp, published July 2, 2026. No source lists public exploit code. No source names a fixed version.

What does this mean?
Severity
8.7 HighCVSS 4.0, GitHub Advisories
Exploit chance
UnscoredNo EPSS score yet
Exploitation
None knownNot on CISA’s or ENISA’s lists
Fixed in
No fix listedNo source names one

Description

Summary

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task and cross-session operations it should not be allowed to (an IDOR / privilege-boundary bypass).

This advisory bundles the audit's Systemic Pattern A findings: F2, F6, F7, F12 (and the duplicate F19).

Affected versions

@grackle-ai/mcp (with @grackle-ai/plugin-core / @grackle-ai/auth) at 0.132.1 and earlier.

Root cause

  • mcp-server.ts:111-127 (createGrpcClients) sets Authorization: Bearer ${apiKey} (the full server key) on every outbound gRPC call.
  • Backend handlers (updateTask, deleteTask, resumeTask, killAgent, getTask in plugin-core) take only the request message — no AuthContext — and act on whatever ID is passed.
  • Therefore scope must be enforced in each MCP tool handler. Some call assertCallerIsAncestor (task_complete, task_start, session_attach, session_send_input); their destructive siblings do not. New tools that forget the check fail open.

F2 — task_update / task_delete / task_resume bypass ancestry (High)

Location: packages/mcp/src/tools/task.ts:226 (task_update), :393 (task_delete), :465 (task_resume).

These accept an arbitrary taskId with only a ROOT_TASK_ID special-case and no assertCallerIsAncestor. The central dispatcher hardcodes its workspace gate to name === "task_show", and these tools' Zod schemas have no workspaceId field (so the blanket injection is stripped). They are not in DEFAULT_SCOPED_MCP_TOOLS but are in the shipped ORCHESTRATOR_MCP_TOOLS / ADMIN_MCP_TOOLS presets.

Exploit: a scoped agent under the orchestrator persona learns a sibling/parent/cross-workspace task ID (via task_list/task_search), then task_delete {taskId: <victim>} (permanently destroys another agent's or the human parent's work) or task_update to flip a sibling to complete/failed or rewrite its dependency DAG and budgets.

F6 — session_kill / session_resume bypass ancestry (Medium)

Location: packages/mcp/src/tools/session.ts:95 (session_resume), :178 (session_kill).

Call client.killAgent / client.resumeAgent directly with no auth context, unlike session_attach/session_send_input. session_kill is in the orchestrator preset. Exploit: cross-session SIGKILL DoS against a sibling agent or the root orchestrator; foreign session resume.

F7 — Workspaceless scoped token fails open to all workspaces (Medium)

Location: packages/mcp/src/mcp-server.ts:565-600.

The per-task cross-workspace check (task_show block) is gated on authContext.workspaceId being truthy. A task created without a workspace (a supported config) mints a scoped token with pid: "" → workspaceId: undefined, so the check is skipped and the injected workspaceId becomes "", which the backend treats as "all workspaces." Exploit: a malicious agent in a workspaceless session calls task_list (no args) → reads every task in every workspace, then task_show {taskId} cross-workspace (title/description/branch/review-notes). schedule_show (in the default scoped allowlist) resolves by ID with no workspace check. Read-only cross-workspace disclosure.

F12 — Scoped-token revocation is dead code (Low)

Location: packages/auth/src/scoped-token.ts:23,142-149.

The consuming check is wired (auth-middleware.ts:90-92), but revokeTask() is never called outside tests — no task-abort/stop flow invokes it — and the backing revokedTasks Map is in-memory only (lost on restart). A compromised agent that exfiltrated its scoped token keeps authenticating for the full 24h TTL regardless of task lifecycle.

Remediation

  • Systemic fix: enforce scope centrally in the CallToolRequest dispatcher (mcp-server.ts) via a per-tool targetTaskIdArg / targetSessionIdArg descriptor so any tool that targets a task/session fails closed unless the caller is an ancestor (or self).
  • Immediately, add assertCallerIsAncestor (or self-or-ancestor) to task_update, task_delete, task_resume, session_kill, session_resume, mirroring task_complete/task_start.
  • F7: do not fail open on empty workspaceId — treat a scoped non-root caller with no workspace as having access to no workspace; apply the task_show membership check whenever the caller is scoped and not ROOT_TASK_ID; add a per-id membership check to schedule_show.
  • F12: wire revokeTask() into task-abort/stop flows with SQLite-backed persistence (like channel-grant revocation), or remove the dead API and document the 24h window.
  • Add regression tests mirroring the existing task_complete ancestor tests for each mutator.

CWEs

CWE-862 (Missing Authorization), CWE-639 (Authorization Bypass Through User-Controlled Key / IDOR), CWE-613 (Insufficient Session Expiration).

Attack profile

How easy it is to exploit

Attack vector
Network
Complexity
Low
Privileges needed
Low
User interaction
None

What an attacker can affect

Confidentiality
High
Integrity
High
Availability
High

Affected packages

Affected packages
PackageAffectedFixed
@grackle-ai/mcp
npm
through 0.132.1None known
@grackle-ai/plugin-core
npm
through 0.132.1None known
@grackle-ai/auth
npm
through 0.132.1None known

What each source reports

GitHub Advisories Score used8.7

Grackle: Fail-open authorization in the MCP tool layer lets scoped agents perform cross-task and cross-session mutations (IDOR)

Summary

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task and cross-session operations it should not be allowed to (an IDOR / privilege-boundary bypass).

This advisory bundles the audit's Systemic Pattern A findings: F2, F6, F7, F12 (and the duplicate F19).

Affected versions

@grackle-ai/mcp (with @grackle-ai/plugin-core / @grackle-ai/auth) at 0.132.1 and earlier.

Root cause

  • mcp-server.ts:111-127 (createGrpcClients) sets Authorization: Bearer ${apiKey} (the full server key) on every outbound gRPC call.
  • Backend handlers (updateTask, deleteTask, resumeTask, killAgent, getTask in plugin-core) take only the request message — no AuthContext — and act on whatever ID is passed.
  • Therefore scope must be enforced in each MCP tool handler. Some call assertCallerIsAncestor (task_complete, task_start, session_attach, session_send_input); their destructive siblings do not. New tools that forget the check fail open.

F2 — task_update / task_delete / task_resume bypass ancestry (High)

Location: packages/mcp/src/tools/task.ts:226 (task_update), :393 (task_delete), :465 (task_resume).

These accept an arbitrary taskId with only a ROOT_TASK_ID special-case and no assertCallerIsAncestor. The central dispatcher hardcodes its workspace gate to name === "task_show", and these tools' Zod schemas have no workspaceId field (so the blanket injection is stripped). They are not in DEFAULT_SCOPED_MCP_TOOLS but are in the shipped ORCHESTRATOR_MCP_TOOLS / ADMIN_MCP_TOOLS presets.

Exploit: a scoped agent under the orchestrator persona learns a sibling/parent/cross-workspace task ID (via task_list/task_search), then task_delete {taskId: <victim>} (permanently destroys another agent's or the human parent's work) or task_update to flip a sibling to complete/failed or rewrite its dependency DAG and budgets.

F6 — session_kill / session_resume bypass ancestry (Medium)

Location: packages/mcp/src/tools/session.ts:95 (session_resume), :178 (session_kill).

Call client.killAgent / client.resumeAgent directly with no auth context, unlike session_attach/session_send_input. session_kill is in the orchestrator preset. Exploit: cross-session SIGKILL DoS against a sibling agent or the root orchestrator; foreign session resume.

F7 — Workspaceless scoped token fails open to all workspaces (Medium)

Location: packages/mcp/src/mcp-server.ts:565-600.

The per-task cross-workspace check (task_show block) is gated on authContext.workspaceId being truthy. A task created without a workspace (a supported config) mints a scoped token with pid: "" → workspaceId: undefined, so the check is skipped and the injected workspaceId becomes "", which the backend treats as "all workspaces." Exploit: a malicious agent in a workspaceless session calls task_list (no args) → reads every task in every workspace, then task_show {taskId} cross-workspace (title/description/branch/review-notes). schedule_show (in the default scoped allowlist) resolves by ID with no workspace check. Read-only cross-workspace disclosure.

F12 — Scoped-token revocation is dead code (Low)

Location: packages/auth/src/scoped-token.ts:23,142-149.

The consuming check is wired (auth-middleware.ts:90-92), but revokeTask() is never called outside tests — no task-abort/stop flow invokes it — and the backing revokedTasks Map is in-memory only (lost on restart). A…

Record
GHSA-f9ff-5x35-7gfw
CVSS 4.0
8.7 HighCVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Weakness
CWE-613, CWE-639, CWE-862
Updated
2026-11-02
OSV8.7

Grackle: Fail-open authorization in the MCP tool layer lets scoped agents perform cross-task and cross-session mutations (IDOR)

Summary

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task and cross-session operations it should not be allowed to (an IDOR / privilege-boundary bypass).

This advisory bundles the audit's Systemic Pattern A findings: F2, F6, F7, F12 (and the duplicate F19).

Affected versions

@grackle-ai/mcp (with @grackle-ai/plugin-core / @grackle-ai/auth) at 0.132.1 and earlier.

Root cause

  • mcp-server.ts:111-127 (createGrpcClients) sets Authorization: Bearer ${apiKey} (the full server key) on every outbound gRPC call.
  • Backend handlers (updateTask, deleteTask, resumeTask, killAgent, getTask in plugin-core) take only the request message — no AuthContext — and act on whatever ID is passed.
  • Therefore scope must be enforced in each MCP tool handler. Some call assertCallerIsAncestor (task_complete, task_start, session_attach, session_send_input); their destructive siblings do not. New tools that forget the check fail open.

F2 — task_update / task_delete / task_resume bypass ancestry (High)

Location: packages/mcp/src/tools/task.ts:226 (task_update), :393 (task_delete), :465 (task_resume).

These accept an arbitrary taskId with only a ROOT_TASK_ID special-case and no assertCallerIsAncestor. The central dispatcher hardcodes its workspace gate to name === "task_show", and these tools' Zod schemas have no workspaceId field (so the blanket injection is stripped). They are not in DEFAULT_SCOPED_MCP_TOOLS but are in the shipped ORCHESTRATOR_MCP_TOOLS / ADMIN_MCP_TOOLS presets.

Exploit: a scoped agent under the orchestrator persona learns a sibling/parent/cross-workspace task ID (via task_list/task_search), then task_delete {taskId: <victim>} (permanently destroys another agent's or the human parent's work) or task_update to flip a sibling to complete/failed or rewrite its dependency DAG and budgets.

F6 — session_kill / session_resume bypass ancestry (Medium)

Location: packages/mcp/src/tools/session.ts:95 (session_resume), :178 (session_kill).

Call client.killAgent / client.resumeAgent directly with no auth context, unlike session_attach/session_send_input. session_kill is in the orchestrator preset. Exploit: cross-session SIGKILL DoS against a sibling agent or the root orchestrator; foreign session resume.

F7 — Workspaceless scoped token fails open to all workspaces (Medium)

Location: packages/mcp/src/mcp-server.ts:565-600.

The per-task cross-workspace check (task_show block) is gated on authContext.workspaceId being truthy. A task created without a workspace (a supported config) mints a scoped token with pid: "" → workspaceId: undefined, so the check is skipped and the injected workspaceId becomes "", which the backend treats as "all workspaces." Exploit: a malicious agent in a workspaceless session calls task_list (no args) → reads every task in every workspace, then task_show {taskId} cross-workspace (title/description/branch/review-notes). schedule_show (in the default scoped allowlist) resolves by ID with no workspace check. Read-only cross-workspace disclosure.

F12 — Scoped-token revocation is dead code (Low)

Location: packages/auth/src/scoped-token.ts:23,142-149.

The consuming check is wired (auth-middleware.ts:90-92), but revokeTask() is never called outside tests — no task-abort/stop flow invokes it — and the backing revokedTasks Map is in-memory only (lost on restart). A…

Record
GHSA-f9ff-5x35-7gfw
CVSS 4.0
8.7 HighCVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Weakness
CWE-613, CWE-639, CWE-862
Updated
2026-11-02

History

History
RecordedChangedSources
first indexedGitHub Advisories, OSV

Dates are when Every CVE recorded each change. Upstream dates may differ.

References

This page doesn't list everything

Left out here: the rest of 2 source descriptions. The JSON record has them.

MoveEnter Open