ellmos-ai

Local AI media editor for video, audio and podcasts with local transcription, transcript-based cuts, Hyperframes motion graphics and agent-driven edits

Model-agnostic computer-use core: one agent loop for Claude, OpenAI CUA, and a mock backend — normalized coordinates, canonical action schema, central safety gate.

Agent-ready project documentation template with START/STATE/TODO/DONE, workflows, lightweight tooling, and LLM-friendly project memory. Independent project - not affiliated with Anthropic, OpenAI or Google.

Local-first AI storyboard director and scene continuity orchestrator bridging video generation, CDP, and multitrack audio

Local-first MCP server for headless Blender asset QA: background script runs and FBX reimport verification. No add-on, no TCP port.

Standalone web scraper (get/links/forms/headers/extract/screenshot), extracted from the BACH system. Python stdlib core + optional extras, with SSRF guard.

Private, evidence-first document agent with persistent local memory, exact citations, reversible actions, and bounded Nemotron reasoning.

Local, application-neutral session checkpoint storage with integrity checks.

Human-readable GUI demonstration recordings for CLI and agent workflows.

Local cron, interval, and scheduled task execution engine with deterministic lease discipline and run receipts

Portable, zero-dependency multi-agent file-lock system — Exclusive and Team Locks (LOCK*.txt) with scopes, expiry, stale-cleanup, cloud-sync support, and a fast overview cache.

Local-first pointer registry and materialization for agent hooks (sibling to policy-registry)

Local-first pointer registry for policies, rules, and decisions using cryptographic SHA-256 source references

Standalone template and minimal runtime kernel for self-contained, offline-runnable agent skills and modules

Neutral domain core for the account/bank-balance domain: bank account CRUD, CAMT balance import, privacy-safe read-only transit projection (wave 1, D-20260903-003)

Local-first MCP server for server operations: HTTP health checks, Apache/Nginx log analysis, dry-run deploy manifests and mail diagnostics

Manifest-driven composition of the local agent-ops ecosystem: ticket routing, locking, cross-machine sync, decision-avatar, skills, and an MCP control plane for CLI coding agents

Alpha MCP server for local-first LLM orchestration: memory, knowledge, state, routing, swarm planning, API discovery, tests, automation, and plugins.

Evidence-based system audits across machines with meta-audit bundling -- four tokens, aggregation ladder, model-manual meta reports

Interactive maps of the ellmos-ai ecosystem: module circuit map, skill library, stack composer