Sample proposal: pentest findings triage with the Darkmoon MCP server (03-integrate/tools)

Aperta
#320 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub

I maintainer di solito rispondono entro 1 giorno

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
3/5
Tempo stimato
1-2 giorni
Idoneità per principianti
25/100
Tipo di issue
Funzionalità
Chiarezza
Abbastanza chiara
Stato di attività
Attiva
Stack tecnologico
nodejs, python
Ambito
ai, security

Direzione di ricerca

This proposes a new sample at python/03-integrate/tools/darkmoon/, modelled on the existing exa and tavily tool integrations and the use-case template. Read those two first, then the MCPClient stdio wiring and the tool_filters allowlist the proposal describes. Work starts only after maintainers confirm the sample fits and the vendor affiliation is acceptable; then the PR should add the README, findings_triage.py, requirements.txt and .env.example. Done means the read-only tools work through the filter and the run_pentest launch stays out of the model's reach.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

Affiliation disclosure up front: I am part of the team building Darkmoon, the project this sample uses. Happy to be told no.

There is no sample yet for putting a security-testing platform behind a Strands agent with a narrow blast radius. I would like to contribute python/03-integrate/tools/darkmoon/, following the use-case template and the structure of the existing tool integrations (exa, tavily).

It connects a Strands MCPClient (stdio) to @darkmoon_ai/mcp-server and uses tool_filters={"allowed": [...]} so the model only gets the read-only tools (list_campaigns, get_findings, get_run_status) and writes a prioritized triage brief. Launching a pentest (run_pentest) is not given to the model: it is a separate scan command with an allowlist and a typed confirmation.

Honest constraint: the MCP server talks to the Darkmoon Dashboard API, which is part of Darkmoon Pro (self-hosted); the engine and CLI are open source (GPL-3.0). The README says this at the top. I verified the MCP wiring, the tool filtering and the gated launch against a local mock of that API with strands-agents 1.57.2; the Bedrock call itself needs the maintainers' or readers' own AWS access.

A working sample (README, findings_triage.py, requirements.txt, .env.example) is ready and I will open it as a PR once you confirm it fits.

Lingua principale
Python
Stelle
897
Fork
450
Merge medio
1g 9h
PR unite (30g)
9

Preparare l'ambiente

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di strands-agents/samples

Tutte le issue di strands-agents/samples

Issue simili

Altre issue su Python

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.