Operator-grade systems, documented in public.
vespid.ai is an engineering surface for products, infrastructure notes, and field reports around systems built for humans and agents. The goal is simple: ship real work, explain it clearly, and leave behind documentation that stays useful.
- Public build logs
- Architecture decisions
- Operator workflows
- Human + agent systems
Less personal-site theater. More durable engineering signal.
This site is meant to become a working memory for shipped systems: project surfaces, operational docs, and notes that remain readable after the launch adrenaline fades.
Project surfaces
Each project starts with a public-facing overview before expanding into a deeper doc tree.
Durable docs
Architecture rationale, setup guides, and operator playbooks live next to the work instead of drifting away from it.
Field notes
Launch notes, postmortems, and stack decisions are written close to implementation, not reconstructed later.
Active repositories under ~/vespid
The homepage now reflects the current vespid repo set: one flagship product, one delegation-security thesis, one portability toolkit, and one lower-level runtime surface.
vespid_voice
Android-first voice control with realtime orchestration, confirmation policy, and agent bridges.
SkillAuth
Reference CLI and execution model for short-lived, scoped agent access to login-gated platforms.
hermes-profile-sync
Export, sanitize, sync, and restore Hermes profiles without treating ~/.hermes as one unsafe blob.
vespid
Execution primitives for local sandboxing, brokered runtime control, session storage, and vault-backed state.
Explore the system
Section titled “Explore the system”What is being built
Current active lanes, their constraints, and the product rationale behind them.
The durable layer
Reference material, architecture notes, and operator-facing documentation designed to age well.
What was learned while shipping
Essays, postmortems, and technical notes worth keeping within reach.
Prefer a crisp project surface over vague promises.
Keep notes and operational knowledge close to the implementation.
Explanations should survive past the launch cycle and help the next build move faster.