Persistent memory + workflow orchestration as an MCP server. 73 canonical tools across 20 namespaces (+3 env-gated extras), SQLite-backed, with an LLM-as-judge drift pipeline that produces artifact-anchored verdicts. Pre-release (alpha.31); Phase 21 wiring in progress. Postgres driver is incomplete.
Skill Claude CodeCodex
Use for governance, memory, drift detection, and audit trail via dark-memory-mcp. Covers 52 canonical + 3 red-team tools across 16 namespaces: session lifecycle, agentmemory CRUD+search, vibe-flow spec/artifact publish + drift, LLM-as-judge, delegation+mindset, research, observability, error observatory, governance…
Skill Claude CodeCodex
Skill "vibe-loop-git" from Opita-Code/dark-memory-mcp, covering vibe-loop-git — 13-loop vibe protocol for dark-memory, what this mod is, what this mod is not, when to load this skill and the 13-loop protocol (in dependency order).
Skill Claude CodeCodex
Skill "claude-code" from Opita-Code/dark-memory-mcp, covering vibe-loop-git — claude-code adapter, claude-code-specific invocation, mount the mod, via the claude-code skills add command) and quick start.
Skill Claude CodeCodex
Skill "claude-desktop" from Opita-Code/dark-memory-mcp, covering vibe-loop-git — claude-desktop adapter, claude-desktop-specific invocation, mount the mod, quick start and claude-desktop env vars.
Skill Claude CodeCodex
OpenAI Codex is OpenAI's coding agent. It supports custom skills via the /.codex/skills/ directory.
Skill Claude CodeCodex
Skill "opencode" from Opita-Code/dark-memory-mcp, covering vibe-loop-git — opencode adapter, opencode-specific invocation, quick start, start a session and invoke the mod.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: