CVE-2026-56838: PraisonAI recipe.run_stream skips dangerous-tool policy enforcement
PraisonAI recipe.runstream() skips dangerous-tool policy enforcement Summary PraisonAI recipe execution blocks default-denied dangerous tools unless the caller explicitly passes allowdangeroustools=True. The normal recipe.run() path enforces this with checktoolpolicy().
Published Jun 18, 2026 · Updated Jul 20, 2026
Affected packages
praisonai
- Ecosystem
- PyPI
- Affected
- >= 4.5.87, <= 4.6.58
- Fixed in
- 4.6.59
- Check
pip show praisonai
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
PraisonAI recipe.run_stream() skips dangerous-tool policy enforcement
Summary
PraisonAI recipe execution blocks default-denied dangerous tools unless the caller explicitly passes allow_dangerous_tools=True. The normal recipe.run() path enforces this with _check_tool_policy(). The streaming path, recipe.run_stream(), loads the same recipe, checks dependencies, and then calls _execute_recipe() without running the dangerous-tool policy check.
As a result, a recipe that honestly declares execute_command in TEMPLATE.yaml requires.tools is denied by recipe.run(), but reaches the execution engine through recipe.run_stream() with allow_dangerous_tools=False.
The local PoV uses a harmless printf canary, explicitly unsets PRAISONAI_AUTO_APPROVE, and avoids network access.
Affected Product
- Repository:
MervinPraison/PraisonAI - Package:
praisonai - Components:
src/praisonai/praisonai/recipe/core.pysrc/praisonai/praisonai/recipe/serve.pysrc/praisonai/praisonai/cli/features/recipe.pysrc/praisonai-agents/praisonaiagents/workflows/yaml_parser.pysrc/praisonai-agents/praisonaiagents/workflows/workflows.py
Validated affected:
- current main
2f9677abb2ea68eab864ee8b6a828fd0141612e1(v4.6.57-4-g2f9677ab) v4.6.57v4.6.56v4.6.10v4.6.9v4.5.128v4.5.120v4.5.96v4.5.87
Suggested affected range: >= 4.5.87, <= 4.6.57.
PyPI lists PraisonAI 4.6.57 as the latest release on 2026-06-13.
Earlier tested tags through v4.5.85 failed in this source checkout before the tested workflow path due an unrelated praisonaiagents.output.models import error. They are not claimed fixed or unaffected.
Root Cause
recipe.run() enforces the dangerous-tool gate:
if not options.get("allow_dangerous_tools", False):
policy_error = _check_tool_policy(recipe_config)
if policy_error:
return RecipeResult(..., status=RecipeStatus.POLICY_DENIED, ...)recipe.run_stream() has a sibling execution path. It loads the recipe and checks dependencies, but then goes directly to execution:
recipe_config = _load_recipe(name, offline=options.get("offline", False))
...
output = _execute_recipe(recipe_config, merged_config, session_id, options)There is no equivalent _check_tool_policy() call in run_stream() before execution or before the dry-run shortcut.
The CLI exposes this path via praisonai recipe run <recipe> --stream, and the recipe HTTP server exposes it as POST /v1/recipes/stream.
Why This Is Not Intended Behavior
The normal recipe path clearly treats declared dangerous tools as denied by default. A control recipe with TEMPLATE.yaml requires.tools: [execute_command] returns:
Tool 'execute_command' is denied by default. Use allow_dangerous_tools=True to override.That operator-facing override should not depend on whether the caller requests streaming output. PraisonAI's own docs describe approval as requiring a human or configured channel before risky tools run, describe security environment variables as opt-in access for dangerous operations with secure defaults, and describe policy controls as blocking dangerous operations.
This is distinct from the prior report PRAI-CAND-011:
PRAI-CAND-011covers workflow tool declarations that are omitted fromTEMPLATE.yaml requires.tools.- This report covers a sibling entrypoint that skips the policy check even when
TEMPLATE.yamlcorrectly declares the dangerous tool.
It is also distinct from the published Recipe-server authentication fail-open advisory. That advisory covers missing authentication secrets. This report assumes the attacker has whatever access is already needed to invoke recipe streaming and focuses on the missing dangerous-tool policy guard.
Local PoV
Run:
python3 poc/pov_prai_cand_012_stream_policy_bypass.pyExpected output includes:
{
"ok": true,
"policy_error": "Tool 'execute_command' is denied by default. Use allow_dangerous_tools=True to override.",
"control_recipe_status": "policy_denied",
"execution_reached": [
{
"recipe": "declared-dangerous-stream",
"declared_required_tools": ["execute_command"],
"allow_dangerous_tools": false
}
],
"workflow_approve_tools": ["execute_command"],
"runner_tool_names": ["execute_command"],
"command_stdout": "PRAI-CAND-012-CANARY",
"operator_env_auto_approve": null
}The PoV creates a temporary recipe that declares execute_command in TEMPLATE.yaml requires.tools.
Control:
recipe.run(..., options={"force": True})returnspolicy_denied.
Bypass:
recipe.run_stream(..., options={"force": True})emits theexecutingevent and reaches_execute_recipe()whileallow_dangerous_toolsremains false.- The same recipe workflow resolves
execute_commandand preservesapprove: [execute_command]. - With the workflow approval context installed, the resolved tool runs the harmless local command
printf PRAI-CAND-012-CANARY.
The PoV monkey-patches _execute_recipe() only to prove that run_stream() crosses the policy boundary without invoking an LLM. The command canary is executed directly through the same resolved workflow tool and approval context to keep the proof deterministic and local-only.
Impact
If an operator runs an untrusted recipe through streaming mode, or exposes the recipe streaming API to users who can choose recipe names or URIs, the recipe can reach execution with default-denied tools even though the caller did not set allow_dangerous_tools=True.
If the workflow reaches the approved execute_command tool call, commands run with the privileges of the PraisonAI process. The exact trigger depends on the workflow and model/tool-call path, but the dangerous-tool policy boundary is already bypassed before execution.
The HTTP recipe sidecar is documented as a localhost REST API with SSE streaming and optional API-key/JWT authentication. This report does not claim default unauthenticated network RCE. In authenticated or exposed sidecar deployments where lower-trust users can invoke /v1/recipes/stream, the same policy gap can become a remote recipe-execution issue.
Suggested Fix
Centralize recipe preflight enforcement so every execution mode uses the same guard:
- Run
_check_tool_policy(recipe_config)inrun_stream()unlessoptions["allow_dangerous_tools"]is true. - Perform that check before both dry-run and real execution, matching
recipe.run(). - Prefer a shared helper for dependency checks, dangerous-tool policy checks, and dry-run handling so future entrypoints cannot drift.
- Add regression tests:
- declared dangerous tool is denied by
recipe.run(); - the same declared dangerous tool is denied by
recipe.run_stream(); allow_dangerous_tools=Truepreserves the intended opt-in behavior;/v1/recipes/streammaps a policy denial to a non-success SSE event or equivalent HTTP failure.
Credited to rexpository.
References
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology
More PyPI vulnerabilities
All PyPI entriesCVE-2026-61438
PraisonAI vulnerable to Remote Code Execution via Broken AST Sandbox
high7.3Oct 29, 2026CVE-2026-60085
PraisonAI: SecurityPolicy command/path/import restrictions are completely unenforced by the default SubprocessSandbox backend
high8.7Oct 8, 2026CVE-2026-60086
PraisonAI: Prompt-injection defense blocks only when 3+ detector families fire simultaneously; realistic single-vector injections pass through unblocked
medium5.3Oct 8, 2026CVE-2026-60088
PraisonAI: Project custom command templates can read outside-workspace files into model prompts
medium5.5Oct 8, 2026CVE-2026-60090
PraisonAI: PGVector and Cassandra knowledge stores interpolate vector dimensions into DDL
mediumOct 8, 2026CVE-2026-60091
PraisonAI: Jobs API is unauthenticated by default and allows attacker-controlled webhook SSRF
high7.2Oct 8, 2026
