How artemfire1980/hermes-titan works
Hermes-Titan is an autonomous engineering and research platform built around an external Hermes Agent, with Telegram as its documented user interface. Its principal workflow is Telegram → Hermes → Aider → Git; a distinct research workflow runs topic research through search, page extraction, evidence analysis, and cited summaries. The graph also preserves the documented task board, memory, skills, and MCP integrations. Research-module connections are based on shown calls or the source index; unsampled integrations are shown only where the README documents them.
Main entry points: runner.py Research orchestrator · aider_runner.py Aider runner · research_runner.py Research CLI · search.py Web search
Components 22
User interaction
- Telegram
- Hermes Agent
Engineering workflow
- Aider runner
scripts/aider_runner.py - Git repository
- Task ledger
scripts/task_ledger.py
Research engine
- Research CLI
scripts/research_runner.py - Research orchestrator
scripts/research/runner.py - Web search
scripts/research/search.py - Page extraction
scripts/research/fetch.py - LLM gateway
scripts/research/llm.py - Evidence analysis
scripts/research/evidence.py - Source scoring
scripts/research/scoring.py - Research checkpoints
scripts/research/checkpoint.py - Cited summaries
scripts/summary_generator.py
Platform capabilities
- Agent skills
skills - Holographic memory
configs/holographic-entities.patch - Kanban board
External integrations
- SearXNG
configs/searxng - FreeLLM API
- MCP services
- Telegram notifications
scripts/send_to_telegram.py
Other components
- User
Connections 21
- User → Telegram: sends request
README.md - Telegram → Hermes Agent: routes request
README.md - Hermes Agent → Aider runner: delegates code
README.md - Aider runner → Task ledger: records tasks
scripts/aider_runner.py - Aider runner → Git repository: changes code
README.md - Hermes Agent → Research CLI: starts research
- Research CLI → Research orchestrator: runs research
scripts/research_runner.py - Research orchestrator → Web search: queries sources
scripts/research/runner.py - Web search → SearXNG: searches web
README.md - Research orchestrator → Page extraction: fetches pages
scripts/research/runner.py - Research orchestrator → LLM gateway: extracts evidence
scripts/research/runner.py - LLM gateway → FreeLLM API: calls API
README.md - Research orchestrator → Evidence analysis: validates claims
scripts/research/runner.py - Research orchestrator → Source scoring: scores sources
scripts/research/runner.py - Research orchestrator → Research checkpoints: saves progress
scripts/research/runner.py - Research orchestrator → Cited summaries: generates summary
scripts/research/runner.py - Hermes Agent → Agent skills: uses skills
README.md - Hermes Agent → Holographic memory: uses memory
README.md - Hermes Agent → Kanban board: manages tasks
README.md - Hermes Agent → MCP services: connects services
README.md - Telegram notifications → Telegram: sends updates
README.md