langflow-darkmoon

A Langflow Extension Bundle (distribution lfx-darkmoon) for Darkmoon, the local, privacy first autonomous AI penetration testing engine (GPLv3). Drop these components into a Langflow flow to trigger a Darkmoon pentest against a target you are authorised to assess, pull back the findings, and review the fix pull requests Darkmoon prepares, over your own self-hosted Darkmoon Dashboard API.

Darkmoon runs and validates security tests. It does not, and these components do not, guarantee that a system is secure. Findings can include false positives and must be reviewed by a qualified human. Only run assessments against systems you own or have explicit written authorisation to test. These components never merge a pull request; every fix is left for a person to review and merge.

Components

The bundle registers under the darkmoon group. Each component is addressable by the canonical namespaced id ext:darkmoon:<Class>@official, for example ext:darkmoon:DarkmoonRunPentest@official.

Install

Darkmoon is self hosted, so you point the components at your own Darkmoon Dashboard API (the FastAPI service shipped with Darkmoon, typically port 8000). There is no public endpoint. Each component logs in with POST /api/v1/auth/login to obtain a short lived JWT.

This is a standalone Langflow Extension Bundle: install it alongside Langflow and it is discovered automatically through the langflow.extensions entry point. No fork of Langflow and no LANGFLOW_COMPONENTS_PATH are required.

pip install langflow lfx-darkmoon
langflow run

Restart the Langflow server; the four Darkmoon components appear in the visual editor under the Darkmoon bundle group. Set the base URL, username and password on a component, and a target on Darkmoon Run Pentest.

Remediation and secrets (Pro)

The Darkmoon dashboard and the remediation feature that prepares fix pull requests are paid Pro capabilities. The open source Darkmoon (the CLI engine) focuses on finding, proving and reporting findings locally. When remediation is enabled it needs a credential reference, an opaque id of a credential stored in Darkmoon's encrypted vault (created in the dashboard), not a raw token. Raw source control secrets never travel through these components, and the remediation agent only ever prepares a pull request for human review; it never merges. Darkmoon List Pull Requests can read those pull requests through the API.

Privacy

Darkmoon keeps assessment work on your own infrastructure and applies a privacy gateway so the language model works over placeholders in place of your real hosts, IPs and credentials. These components send data only to the base URL you configure.

Darkmoon at a glance

Darkmoon is a GPLv3, self hosted autonomous AI pentester: an LLM orchestrates specialist agents and offensive tools over MCP, runs on a local model, and proves each finding with a real exploit. See ASCIT31/Dark-Moon.

Layout

langflow-darkmoon/
├── pyproject.toml                 # distribution: lfx-darkmoon
├── src/
│   └── lfx_darkmoon/
│       ├── __init__.py            # re-exports the four components
│       ├── extension.json         # Extension Bundle manifest (v1)
│       └── components/
│           └── darkmoon/          # the bundle (path in extension.json)
│               ├── run_pentest.py
│               ├── get_findings.py
│               ├── list_campaigns.py
│               ├── list_pull_requests.py
│               ├── _base.py          # shared inputs + client mixin (not a Component)
│               ├── darkmoon_client.py# dependency free, transport injected API client
│               └── transport.py      # requests transport + client factory
└── tests/

Development

pip install langflow          # brings lfx (the Bundle API surface)
pip install -e ".[test]"      # install this bundle editable
lfx extension validate src/lfx_darkmoon   # static manifest + bundle check
lfx extension list            # confirm it is discovered at @official
pytest -q

The API client (darkmoon_client.py) is dependency free and transport injected, so its logic is unit tested without a network (mock transport). The component tests instantiate the real Langflow (lfx) components, inject a mock transport, run each output, and assert each builds a valid Langflow frontend node. A full end to end run additionally requires a running Darkmoon instance pointed at an authorised target.

Intra bundle imports are absolute (from lfx_darkmoon.components.darkmoon...), as required by the Langflow bundle loader, which does not support relative imports between bundle modules.

Related integrations

Darkmoon also ships integrations for n8n, Activepieces, Dify, LangChain and GitHub Actions.

License

MIT, see LICENSE. Not affiliated with Langflow; "Langflow" is a trademark of its respective owner.