Repository navigation
DarkMoon v1.7.0 — AGPL-3.0, executive security views, Terminal Explorer, attack paths
First DarkMoon Community release under the GNU AGPL v3.0. Everything up to commit edfe9cb (27 Oct 2026), including v1.6.0, stays under GPL-3.0, and so do the earlier outside contributions listed in LICENSING.md. LICENSING.md and the CLA are drafts still under legal review (this release is published on a provisional validation); their wording may change, the licence text itself is the official GNU AGPL v3.0.
📊 Executive security views (shared deterministic core)
- Cyber Score v1: deterministic, coverage-gated (not rated / limited / rated), published caps and methodology (docs/cyber-score-methodology.md). It rates what one campaign found in its tested scope; it is not a security level or a compliance statement.
- Prioritized Treatment Plan: one action per asset × weakness class, indicative effort, statuses with history, retest-based verification with a coverage gate, risk acceptance with a named owner and an expiry.
- Frameworks: ISO 27001 / GDPR art. 32 / NIS2 art. 21(2) / AI Act art. 15 in four states (weakness observed, none observed in tested scope, not tested, not technically assessable) — counts, never a percentage or « compliant ».
- Application profile, maintenance windows (Community warns; the decision is shown in the report), security event taxonomy, report model v2 rendered to Markdown.
- Coverage signals recorded server-side; a terminal campaign status can no longer be re-finalised; agent labels cannot inflate the score.
🖥️ Terminal Explorer (OpenCode plugin)
Workspace with the executive block, projects, campaigns, findings, treatment plan, attack paths you can walk step by step, frameworks, profile, legend — keyboard-driven, readable at 80 columns, scrollable legend. Screenshots: docs/pics/terminal-explorer.
🧭 Attack paths (architecture D)
Agents declare what enabled what (enabled_by); every link stays a hypothesis unless evidence supports it. Community shows Possible attack paths only (it keeps no execution ledger). Honest limits and experiments: darkmoon-attack-paths-research.
🔒 Privacy & safety
- Tool-output filter fails closed; private keys and table secrets masked before the model;
/proc/nethex addresses masked. - Target-safety rule in every sub-agent; no state-changing or self-hanging probes.
dashboard_*writes bound to the calling run when a run key is present.
📚 Docs & research
docs/full.mdupdated (architecture, MCP tools — 13 in Community —, attack paths, executive security, limits).- Two public research repositories: darkmoon-attack-paths-research and darkmoon-visualization-research; research series on the blog.
As always, DarkMoon reports safe metadata only — severity, status, MITRE, CVE ids, evidence descriptions — never exploits, secrets or tokens.