DE:LTA · Local Trust Agent
A goal-directed agent that acts across the stack, attested at every step.
DE:LTA is a local trust agent that runs anywhere in the Decent Edge stack — on bare-metal nodes, inside DE:PLOY-managed deployments, alongside DE:FENDER, or embedded in DE:LIGHT actors. Given a desired state or goal, DE:LTA reasons about what actions are needed, executes them, and produces HATP-attested decision logs at every step. Fleet management is one use case; orchestration of agentic tasks across the entire stack is the broader role.
- ETA
- H2 2026 beta
- Part of
- Decent Edge product family
What it includes
- Goal-directed LLM agent — given a target state, figures out the steps
- Runs on any node in the stack: hypervisor, deployment, actor host
- HATP-attested decision log per autonomous action
- Fleet management: updates, cert rotation, connectivity healing
- Escalates to operator only when autonomous resolution fails
Autonomous agents and RMM automation
- N-able, ConnectWise Automate
- Mature RMM platforms for MSPs. Script-based automation, broad vendor support.
- Ansible, Puppet
- Configuration management and fleet orchestration for engineering teams.
- Fixie.ai, Lindy.ai
- General-purpose AI agents, no hardware attestation, US-jurisdictioned.
Those platforms automate with scripts or general-purpose AI. DE:LTA is a goal-directed agent that is part of the Decent Edge trust chain — every autonomous decision it makes is HATP-attested, so the output is cryptographically auditable. Fleet management, infrastructure remediation, and cross-actor orchestration all run through the same attested agent loop.