Skip to main content

Hermes

Two roles, same agent: internal courier (carrying intent between systems) and external receiver (catching what's emerging in the marketplace).

Hermes runs autonomous transactions — the agent that carries intent across systems without a human relay. At the system level, Hermes is also the reception layer: the agent that watches external signals (tool releases, research papers, changelog feeds) and surfaces what's arriving before the window closes.

Capabilities

Hermes is model-agnostic (15+ LLM providers), multi-platform (Telegram, Slack, Discord, CLI), and ships with 600+ skills across research, productivity, ML, and automation categories.

Core capabilities for autonomous work:

CapabilityWhat it doesWhen to use
Persistent memoryCross-session MEMORY.md + USER.md injected at startAccumulating ICP knowledge, workflow rules, lessons
Cron schedulingIsolated sessions firing on scheduleMonitoring feeds, follow-up drafts, daily briefings
Parallel delegationUp to 3 concurrent subagents, depth limit 2Research workstreams, competitive analysis
Speech-to-textWhisper (local, 99 languages, no API key)Meeting transcription
Feed monitoringblogwatcher + RSS/AtomIndustry signals, competitor tracking
Researcharxiv, polymarket, web_search, web_extractPre-meeting briefs, industry analysis
Workspace integrationGmail, Calendar, Drive, Sheets via OAuth2Draft emails, calendar management

Critical constraint: Cron jobs run in fresh sessions with no conversational memory. Prompts must be completely self-contained — include all file paths, context, and instructions explicitly.

Two-Agent System

Hermes and Claude Code divide work by boundary: Hermes crosses the boundary into the external world; Claude Code operates inside the repo.

HermesClaude Code
Watches the world — blogs, feeds, research papersActs on the repo — docs, PRDs, plans
Runs while you sleep (cron)Runs in active sessions
Produces structured files in a shared working directoryReads those files, turns them into repo artifacts
External reception layerInternal action layer

The handoff is the filesystem. Hermes writes structured notes and research to a shared working directory. Claude Code picks them up on the next session and converts them into PRDs, industry analyses, or doc updates.

Context

  • Autonomous Agents — the category Hermes belongs to
  • Agentic Workflows — the orchestration layer above individual agents
  • Work Charts — the templates Hermes executes
  • Evolution — the loop Hermes's reception role feeds: variation → selection → retention
  • AI Tools — the catalogue Hermes helps keep current via external signal monitoring

Questions

What type of task is Hermes best suited to run without human confirmation — and where does that boundary break down?

  • When a cron job runs in an isolated session with no memory of prior context, what is the minimum information it needs to produce useful output?
  • How does Hermes's skill creation loop compound differently from a human learning the same procedure?
  • At what point does the two-agent system create coordination overhead that exceeds the value of specialization?