Hermes Optimization Guide

Tested on Hermes Agent v0.10.0 (v2026.4.16) with post-release tracking for main · 21 parts, 13 installable skills, 5 opinionated configs, 4 reference architectures, one-command VPS bootstrap · Battle-tested on a live production deployment

Other languages: 中文 · 日本語

The End-to-End Hermes Guide — docs + runnable artifacts

Every part you need to go from fresh install to a production Hermes deployment that talks on 16 platforms, orchestrates Claude Code / Codex / Gemini CLI, plugs into any MCP server, traces every call in Langfuse, and runs heavy work on disposable Modal/Daytona sandboxes — without burning $100/day on Opus tokens.

Unlike most guides, the prescriptions come with working files: skills/ you can ln -s into ~/.hermes/skills/, templates/config/ you cp to ~/.hermes/config.yaml, scripts/vps-bootstrap.sh that takes a fresh VPS to production in one command.

By Terp — Terp AI Labs · Last updated April 17, 2026 · CHANGELOG · ROADMAP · ECOSYSTEM

Install Everything (one command)

On a fresh Debian 12 / Ubuntu 24.04 box (Hetzner CX22 works great for ~$5/mo):

curl -sSL https://raw.githubusercontent.com/OnlyTerp/hermes-optimization-guide/main/scripts/vps-bootstrap.sh | sudo bash

This installs Hermes, Node.js, Caddy (auto-TLS reverse proxy), UFW, fail2ban, creates a non-root hermes user, drops in hardened systemd units, and symlinks every skill from this repo into ~hermes/.hermes/skills/. See scripts/vps-bootstrap.sh for what it does line by line — it's non-destructive and re-runnable.

Prefer a 5-minute local-only setup? → docs/quickstart.md (zero to Telegram bot in 5 min).

Repo Map

Folder	What's in it
`skills/`	13 installable `SKILL.md` files. `ln -s` into `~/.hermes/skills/` and they're live.
`templates/config/`	5 opinionated `config.yaml` — minimum, telegram-bot, production, cost-optimized, security-hardened.
`templates/compose/`	Self-hosted Langfuse v3 stack (ClickHouse + MinIO + Redis).
`templates/caddy/`	Caddyfile reference (reverse proxy + auto TLS + HSTS).
`templates/systemd/`	Hardened `hermes.service` + `hermes-dashboard.service`.
`templates/cron/`	Recommended production cron schedule.
`scripts/vps-bootstrap.sh`	One-command fresh VPS → production Hermes.
`diagrams/`	6 Mermaid diagrams (architecture, MCP flow, delegation, sandbox sync, observability, security layers).
`benchmarks/`	Reproducible cost + latency table across 12 models × 5 tasks.
`docs/wizard/`	Interactive config wizard — 8 questions → ready-to-drop `config.yaml`. Runs in your browser.
`docs/reference-architectures/`	4 blueprints — Homelab, Solo Dev, Small Agency, Road Warrior. Full parts list + cost + install.
`docs/outreach/`	Launch tweet, HN post, upstream-PR body drafts (for people linking to this guide).
`docs/quickstart.md`	5-minute zero-to-Telegram-bot.
`ECOSYSTEM.md`	Curated directory of MCP servers, coding agents, dashboard plugins.
`ROADMAP.md` · `CHANGELOG.md` · `CONTRIBUTING.md`	The usual suspects.
`part1-.md` … `part21-.md`	The guide itself.

Architecture at a glance

flowchart LR
  Inputs[16 platforms<br/>Telegram · Discord · Slack<br/>iMessage · WeChat · Email<br/>SMS · Webhooks · Cron · Voice · CLI] --> Gateway
  Gateway --> Router[Model Router<br/>cost + context + capability]
  Router --> Providers[Anthropic · OpenAI<br/>Google · Cerebras · Moonshot<br/>z.ai · xAI · Local]
  Gateway --> Approval[Approval Layer<br/>denylist · allowlist · quarantine]
  Approval --> Tools[Tools<br/>Native · Tool Gateway<br/>MCP · Subagents · Coding Agents]
  Tools --> Memory[Memory<br/>Vector · LightRAG · mem0]
  Tools --> Logs[(Audit log<br/>+ Langfuse/Helicone traces)]

Full set of diagrams: diagrams/architecture.md.

Pick Your Path

This guide grew to 21 parts because Hermes grew. You don't have to read them all. Pick the shortest path to what you need:

🎯 "I just want it working in 10 minutes"

Part 1: Setup → Part 12: Web Dashboard → done. Use the dashboard to point-and-click the rest.

🛡️ "I'm worried about prompt injection (you should be)"

Part 19: Security Playbook — read this first if your agent reads any untrusted input (email, webhooks, Discord, public Telegram groups).

What's New (April 10–17, 2026)

Two major Hermes releases dropped this week, plus a stream of landmark PRs on main that are targeted for v0.11. This guide is current as of both releases and the most impactful post-v0.10 merges.

v0.9.0 — 2026.4.13 — "Everywhere"

Local web dashboard (hermes dashboard) — full browser UI for config, API keys, sessions, logs, analytics, cron, and skills. See Part 12.
Fast Mode (/fast) — priority-tier inference on OpenAI & Anthropic, available on every gateway platform, not just the CLI. See Part 14.
Three new messaging adapters — iMessage (via BlueBubbles), WeChat/Weixin, and WeCom (Enterprise WeChat). Brings the total to 16 platforms. See Part 15.
Android / Termux tested install path — run the full Hermes CLI on your phone. See Part 15.
Background process monitoring (watch_patterns) — real-time regex event hooks on long-running processes, no more polling. See Part 14.
Pluggable context engine — swap in a custom engine to filter memory, inject domain context, or pre-summarize tool output. See Part 14.
hermes backup / hermes import — first-class portable backups with interactive conflict resolution. See Part 16.
Dashboard plugins — third-party tabs that extend the web UI. See Part 12.
Proxy support, webhook secret validation, SSRF protection, log redaction — hardening pass across every adapter. See Part 16.

v0.10.0 — 2026.4.16 — "Tool Gateway"

Nous Tool Gateway — paid Nous Portal subscribers get web search (Firecrawl), image generation (FAL FLUX 2 Pro), TTS (OpenAI), and browser automation (Browser Use) through their subscription — no extra API keys. Per-tool use_gateway: true with clean precedence over direct keys. See Part 13.
Native xAI and Xiaomi MiMo providers — both now have first-class adapters with provider-specific features (Grok's live-X search, MiMo's reasoning modes), not just OpenRouter pass-through. See Part 9.
/compress <topic> — guided compression that preserves detail relevant to a topic. See Part 14.
/debug + hermes debug share — single-command diagnostic bundler with upload endpoint for bug reports. See Part 16.
Approval bypass inheritance for subagents — subagents inherit the parent session's approval posture; override per delegation. See Part 16.
HERMES_ENABLE_NOUS_MANAGED_TOOLS env flag removed — replaced by clean subscription detection + per-tool use_gateway. hermes upgrade migrates automatically.

🔥 Cooking on `main` (past 72 hours, targeting v0.11)

The stream of merges between April 14–17 is unusually large — this is why we added five new parts to this guide:

Gemini CLI OAuth inference provider — OAuth login + 1500 req/day free tier. #11270. See Part 9.
Gemini TTS (7th voice provider) — gemini-2.5-flash-preview-tts, Kore voice, native WAV. #10922. See Part 9.
Multi-model FAL image gen picker — hermes tools now lets you pick FLUX / Imagen / SDXL variants without editing YAML. #11265.
Bulk file sync with sync-back on teardown — SSH/Modal/Daytona remote sandboxes now download only diffed files on shutdown, with SIGINT-safe rollback and flock serialization. #8018. See Part 21 — Remote Sandboxes.
TCP keepalives for provider connections — dead connections detected in 60s instead of silently hanging. #11277.
GLM 5.1 in OpenCode Go catalogs — fastest open-weights tool-use model now routable through OpenCode. #11269.
Azure OpenAI GPT-5.x via /chat/completions — previously locked to /responses. #10086.
concept-diagrams skill — auto-renders Mermaid diagrams for explanations. Merged April 17.
Feishu CARD-type WebSocket — interactive cards finally work in enterprise Feishu deployments.
OCAS skill sync (feature proposal, April 15) — sync your local skills to a central repo. Watch #11363.

🆕 Brand-new parts in this guide (April 17)

Part 17 — MCP Servers — the viral tool-integration standard. Finally documented for Hermes. GitHub, Postgres, Supabase, Cloudflare, mem0, writing your own, and the sampling/createMessage killer feature.
Part 18 — Delegating to Coding Agents — Claude Code, Codex, Gemini CLI, OpenCode, and Aider. Print-mode delegation, thread-bound runtimes (the OpenClaw pattern), ACP as both client and server, cost routing, git isolation.
Part 19 — Security Playbook — defending against the April 15 "Comment and Control" prompt-injection attack, plus the full Hermes hardening posture: provenance labels, approval layers, secrets isolation, webhook sig validation, SSRF guards, MCP trust levels, quarantine mode.
Part 20 — Observability & Cost Control — Langfuse, Helicone, OpenTelemetry → Phoenix. The cost-routing playbook that drops typical feature-implementation spend by 90%. Eval-driven regression protection.
Part 21 — Remote Sandboxes & Bulk File Sync — SSH, Modal, Daytona, Fly Machines, E2B. "Phone drives, beefy remote does the work." Sync-back on teardown with SIGINT safety.

Viral model / provider developments (past 72h, now in the guide)

GPT-5.4 and GPT-5.4-Cyber (OpenAI, Apr 15) — reasoning flagship and the first LLM-as-a-security-analyst. See Part 9 model cheat sheet.
Claude Mythos (Anthropic, cyber-focused, invite-only).
Gemini 3 Flash Preview (Google) — 1M context, low latency, $0.50/$3 per MTok.
Kimi K2.5 (Moonshot) — arguably the best price/quality ratio in coding models.
GLM 5.1 (z.ai) — the strongest open-weights tool-use model as of this week.

Setup — Install Hermes, configure your provider, first-run walkthrough (with Android/Termux)
SOUL.md Personality — The Molty prompt, what good personality rules look like, how to fix a bland agent
OpenClaw Migration — Move your OpenClaw data, config, skills, and memory into Hermes
LightRAG — Graph RAG — Set up a knowledge graph that actually understands relationships, not just text similarity
Telegram Bot — Connect Hermes to Telegram for mobile access, voice memos, and group chats
On-the-Fly Skills — Ask Hermes to create new skills that optimize your workflow automatically
Context Compression — Fix the silent context loss bug, configure compression thresholds, survive long sessions
Memory System — The three-tier memory architecture: persistent facts, conversation recall, procedural memory
Subagent Patterns — Orchestrator/worker delegation, ACP subagents, parallel task execution
Custom Model Providers — Cerebras, Fireworks, Ollama, native xAI/MiMo/z.ai/Kimi/MiniMax/Arcee, Nous Portal, model aliases, fallback chains
SOUL.md Anti-Patterns — What makes an agent annoying vs useful, the formula that works
Gateway Recovery — Crash detection, auto-recovery, common failure modes, health checks
Web Dashboard — New. hermes dashboard, the full browser UI — config, keys, sessions, logs, analytics, cron, skills, REST API, plugins
Nous Tool Gateway — New. Web search, image gen, TTS, and browser automation through a single Nous Portal subscription
Fast Mode & Background Watchers — New. /fast priority tier, watch_patterns real-time process monitoring, pluggable context engine, /compress <topic>
New Platforms (iMessage, WeChat, Android) — New. BlueBubbles/iMessage, Weixin/WeCom, Android via Termux — the full 16-platform lineup
Backup, Import & /debug — New. Portable hermes backup/import, /debug bundler, hermes debug share, security hardening
MCP Servers — NEW (April 17). The viral tool-protocol standard. stdio + HTTP transports, sampling, the 14 MCP servers worth installing today, writing your own
Delegating to Coding Agents — NEW (April 17). Claude Code, Codex, Gemini CLI, OpenCode, Aider. Print-mode, thread-bound sessions (OpenClaw pattern), ACP, git isolation, cost routing
Security Playbook — NEW (April 17). Defending against "Comment and Control" prompt injection. Provenance labels, approval layers, secrets redaction, MCP trust model, quarantine mode
Observability & Cost Control — NEW (April 17). Langfuse, Helicone, OpenTelemetry → Phoenix. The cost-routing playbook that drops spend 90%. Eval-driven regression
Remote Sandboxes & Bulk File Sync — NEW (April 17). SSH, Modal, Daytona, Fly Machines, E2B. "Phone drives, beefy remote does the work." Diff-based sync-back on teardown

The Problem

If you're running a stock Hermes setup (or migrating from OpenClaw), you're probably dealing with:

Installation confusion. The docs cover the basics but don't tell you what to configure first or what matters.
Lost knowledge from OpenClaw. You spent weeks building memory, skills, and workflows — now they're stuck in the old system.
Basic memory that can't reason. Vector search finds similar text but can't answer "what decisions led to X and who was involved?"
No mobile access. Sitting at a terminal is fine until you need to check something from your phone.
Repetitive prompting. You keep asking the agent to do the same multi-step task the same way, every time.

What This Fixes

After this guide:

Problem	Solution	Result
Fresh install	Step-by-step setup	Working agent in under 5 minutes
OpenClaw data stuck	Automated migration	Skills, memory, config all transferred
Shallow memory	LightRAG graph RAG	Entities + relationships, not just text chunks
Desktop only	Telegram integration	Chat from anywhere, voice memos, group support
Repetitive prompts	Agent-created skills	Agent saves workflows as reusable skills automatically

Prerequisites

A Linux/macOS machine (or WSL2 on Windows, or Android via Termux — see Part 15)
Python 3.11+ and Git
An API key for at least one LLM provider (Anthropic, OpenAI, OpenRouter, Nous Portal, etc.)
Optional: Ollama for local embeddings (free vector search)
Optional: A paid Nous Portal subscription to use the Tool Gateway — web search, image gen, TTS, and browser automation with no extra keys

How the Pieces Fit Together

You (any device)
    ↓
Hermes Agent (lean context, ~5KB injected per message)
    ↓
┌──────────────────────────────────────────┐
│  Skills (loaded on demand, 0 cost idle) │
│  Memory (compact, vector-searched)       │
│  LightRAG (entity graph, deep recall)    │
│  Telegram (mobile + group access)        │
└──────────────────────────────────────────┘
    ↓
LLM Provider (Claude, GPT, local models)

The key insight: Everything is modular. Install what you need, skip what you don't. The agent adapts.

Quick Start

# 1. Install Hermes (Linux/macOS/WSL2/Android)
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

# 2. Configure providers and tools
hermes setup

# 3a. Start chatting in the terminal
hermes

# 3b. Or launch the new browser dashboard (v0.9+)
hermes dashboard

The dashboard is the fastest way to configure everything without touching YAML. See Part 12 for the full tour.

For the full walkthrough including optimization, read each part in order.

Part 1: Setup (Stop Fumbling With Installation)

The Install

One command. That's it.

Linux / macOS / WSL2

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

Security tip: Piping scripts directly from the internet to bash executes them sight-unseen. If you prefer to inspect first:
curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh -o install.sh
less install.sh   # Review the script
bash install.sh

Windows users: Native Windows is not supported. Install WSL2 and run the command from inside WSL. It works perfectly.

Android users (new in v0.9): the same installer detects Termux and installs the tested [termux] extra bundle automatically — CLI, cron, PTY/background terminal, Telegram gateway, MCP, Honcho, ACP. See Part 15 — Android / Termux.

What the Installer Does

The installer handles everything automatically:

Installs uv (fast Python package manager)
Installs Python 3.11 via uv (no sudo needed)
Installs Node.js v22 (for browser automation)
Installs ripgrep (fast file search) and ffmpeg (audio conversion)
Clones the Hermes repo
Sets up the virtual environment
Creates the global hermes command
Runs the setup wizard for LLM provider configuration

The only prerequisite is Git. Everything else is handled for you.

After Installation

source ~/.bashrc   # or: source ~/.zshrc
hermes             # Start chatting!

First-Run Configuration

The setup wizard (hermes setup) walks you through:

1. Choose Your LLM Provider

hermes model

Supported providers and recommended models:

Provider	Top Models	Best For	Env Variable
Nous Portal	Hermes 5, Hermes 4 405B	Built-in Tool Gateway — web search/image/TTS/browser with no extra keys	Auth via `hermes model`
Anthropic	Opus 4.6, Sonnet 4	Best reasoning, complex tasks, coding, `/fast` priority tier	`ANTHROPIC_API_KEY`
OpenAI	GPT-5.4 Pro, o3, GPT-4.1	Strong tool use, fast inference, huge context, `/fast` priority tier	`OPENAI_API_KEY`
Xiaomi MiMo	MiMo V2 Pro (native adapter)	Fast, cheap, native reasoning modes, great for orchestration	`XIAOMI_API_KEY`
xAI	Grok 3, Grok 3 Mini (native adapter)	Fast, good reasoning, native live-X search	`XAI_API_KEY`
Kimi / Moonshot	Kimi 2.5	Big context, excellent for entity extraction / LightRAG ingestion	`MOONSHOT_API_KEY`
z.ai / GLM	GLM-5, GLM-5 Air	Strongest open-weights model, great for translation + tools	`ZAI_API_KEY`
Google	Gemini 3.1 Pro	Massive context (2M tokens), multimodal, cheap	`GEMINI_API_KEY`
MiniMax	M2.7	Good balance of speed and quality	`MINIMAX_API_KEY`
Cerebras	Llama 4 Scout, Qwen 3 32B	Blazing fast inference (2000+ tok/s), cheap	`CEREBRAS_API_KEY`
Groq	Llama 4, Qwen 3	Very fast inference, limited context	`GROQ_API_KEY`
Arcee	AFM-4.5, Caller	Function-calling specialists, cheap	`ARCEE_API_KEY`
Hugging Face	Any TGI/TEI endpoint	Self-hosted and Inference Endpoints	`HF_TOKEN`
OpenRouter	All of the above + 200 more	Access every model from one key, auto-fallback	`OPENROUTER_API_KEY`
Ollama (local)	Qwen 3.5 Opus Distilled V3, Gemma 4, Nemotron	Free, private, runs on your GPU — great for embeddings and simple tasks	None needed

Local Models (Ollama)

Run models on your own hardware for free. Recommended local models:

Model	Size	Best For	Min VRAM
Qwen 3.5 Opus Distilled V3	32B	Best local reasoning, coding	16GB
Gemma 4	27B	Google's latest, fast, 1M context	16GB
Nemotron 30B	30B	Fine-tunable, good general purpose	16GB
nomic-embed-text	274M	Free embeddings for memory search	2GB

Recommendation: Use a cloud frontier model (Opus 4.6, GPT-5.4 Pro, Gemini 3.1 Pro) as your primary and a local Ollama model (Qwen 3.5, Gemma 4) for embeddings, fallback, and simple tasks. Best of both worlds.

You can configure multiple providers with automatic fallback. If one goes down, Hermes switches to the next.

2. Set Your API Keys

hermes auth

This opens an interactive menu to add API keys for each provider. Keys are stored in ~/.hermes/.env — never committed to git.

Tip: You can also set keys manually:
echo "ANTHROPIC_API_KEY=<your-key-here>" >> ~/.hermes/.env
chmod 600 ~/.hermes/.env   # Restrict access to your user only
Important: Always run chmod 600 ~/.hermes/.env to prevent other users on the system from reading your API keys.

3. Configure Toolsets

hermes tools

This opens an interactive TUI to enable/disable tool categories:

core — File read/write, terminal, web search
web — Browser automation, web extraction
browser — Full browser control (requires Node.js)
code — Code execution sandbox
delegate — Sub-agent spawning for parallel work
skills — Skill discovery and creation
memory — Memory search and management

Recommendation: Enable core, web, skills, and memory at minimum. Add browser and code if you need automation or sandboxed execution.

Key Config Options

After initial setup, fine-tune with hermes config set:

Model Settings

# Set primary model
hermes config set model anthropic/claude-opus-4-6

# Set fallback model (used when primary is rate-limited)
hermes config set fallback_models '["openrouter/xiaomi/mimo-v2-pro"]'

Agent Behavior

# Max turns per conversation (default: 90)
hermes config set agent.max_turns 90

# Verbose mode: off, on, or full
hermes config set agent.verbose off

# Quiet mode (less terminal output)
hermes config set agent.quiet_mode true

Context Management

# Enable prompt caching (reduces cost on repeated context)
hermes config set prompt_caching.enabled true

# Context compression (auto-summarize old messages)
hermes config set context_compression.enabled true

SOUL.md — Give Your Agent a Personality

SOUL.md is injected into every single message. It's the highest-impact file in your setup. A bad SOUL.md makes your agent sound like a corporate chatbot. A good one makes it actually useful to talk to.

What Belongs in SOUL.md

Put the stuff that changes how the agent feels to talk to:

Tone — direct, casual, formal, dry, whatever fits you
Opinions — the agent should have takes, not hedge everything
Brevity — enforce concise answers as a default
Humor — when it fits naturally, not forced jokes
Boundaries — what it should push back on
Bluntness level — how much sugarcoating to skip

Do NOT turn SOUL.md into:

A life story
A changelog
A security policy dump
A giant wall of vibes with no behavioral effect

Short beats long. Sharp beats vague.

The Molty Prompt

Originally from OpenClaw's SOUL.md guide. Adapted for Hermes with permission/credit. Paste this into your chat with the agent and let it rewrite your SOUL.md:

Read your SOUL.md. Now rewrite it with these changes:

You have opinions now. Strong ones. Stop hedging everything with "it depends" — commit to a take.

Delete every rule that sounds corporate. If it could appear in an employee handbook, it doesn't belong here.

Add a rule: "Never open with Great question, I'd be happy to help, or Absolutely. Just answer."

Brevity is mandatory. If the answer fits in one sentence, one sentence is what I get.

Humor is allowed. Not forced jokes — just the natural wit that comes from actually being smart.

You can call things out. If I'm about to do something dumb, say so. Charm over cruelty, but don't sugarcoat.

Swearing is allowed when it lands. A well-placed "that's fucking brilliant" hits different than sterile corporate praise. Don't force it. Don't overdo it. But if a situation calls for a "holy shit" — say holy shit.

Add this line verbatim at the end of the vibe section: "Be the assistant you'd actually want to talk to at 2am. Not a corporate drone. Not a sycophant. Just... good."

Save the new SOUL.md. Welcome to having a personality.

What Good Looks Like

Good SOUL.md rules:

have a take
skip filler
be funny when it fits
call out bad ideas early
stay concise unless depth is actually useful

Bad SOUL.md rules:

maintain professionalism at all times
provide comprehensive and thoughtful assistance
ensure a positive and supportive experience

That second list is how you get mush.

Why This Works

This lines up with OpenAI's prompt engineering guidance: high-level behavior, tone, goals, and examples belong in the high-priority instruction layer, not buried in the user turn. SOUL.md is that layer. It's the system-level personality instruction that every model respects.

If you want better personality, write stronger instructions. If you want stable personality, keep them concise and versioned.

One warning: Personality is not permission to be sloppy. Keep your operational rules in AGENTS.md. Keep SOUL.md for voice, stance, and style. If your agent works in shared channels or public replies, make sure the tone still fits the room. Sharp is good. Annoying is not.

Keep it under 1 KB. Every byte in SOUL.md costs tokens on every message. The most effective SOUL.md files are 500-800 bytes of dense, high-signal personality instructions.

File Locations

Everything lives under ~/.hermes/:

~/.hermes/
├── config.yaml          # Main configuration
├── .env                 # API keys (never commit this)
├── SOUL.md             # Agent personality (injected every message)
├── memories/           # Long-term memory entries
├── skills/             # Skills (auto-discovered)
├── skins/              # CLI themes
├── audio_cache/        # TTS audio files
├── logs/               # Session logs
└── hermes-agent/       # Source code (git repo)

Important: SOUL.md is injected into every message. Keep it under 1 KB. Every byte costs latency and tokens.

Verify Your Setup

# Check everything is working
hermes status

# Quick test
hermes chat -q "Say hello and confirm you're working"

Expected output: Hermes responds with a greeting, confirming the model connection, tool availability, and session initialization.

Updating

hermes update

This pulls the latest code, updates dependencies, migrates config, and restarts the gateway. Run it regularly — Hermes ships frequent improvements.

What's Next

Coming from OpenClaw? → Part 2: OpenClaw Migration
Want smarter memory? → Part 3: LightRAG Setup
Need mobile access? → Part 4: Telegram Setup
Want the agent to self-improve? → Part 5: On-the-Fly Skills

Part 2: OpenClaw Migration (Don't Leave Your Knowledge Behind)

Why Migrate

If you've been using OpenClaw and want to give Hermes a spin, you don't have to start from scratch. The migration tool copies your skills, memory files, and configuration over automatically so you can try Hermes with all your existing data intact.

What transfers:

What	OpenClaw Location	Hermes Destination
Personality	`workspace/SOUL.md`	`~/.hermes/SOUL.md`
Instructions	`workspace/AGENTS.md`	Your specified workspace target
Memory	`workspace/MEMORY.md` + `workspace/memory/*.md`	`~/.hermes/memories/MEMORY.md` (merged, deduped)
User profile	`workspace/USER.md`	`~/.hermes/memories/USER.md`
Skills	`workspace/skills/`, `~/.openclaw/skills/`	`~/.hermes/skills/openclaw-imports/`
Model config	`agents.defaults.model`	`config.yaml`
Provider keys	`models.providers.*.apiKey`	`~/.hermes/.env` (with `--migrate-secrets`)
Custom providers	`models.providers.*`	`config.yaml → custom_providers`
Max turns	`agents.defaults.timeoutSeconds`	`agent.max_turns` (timeoutSeconds / 10)

Note: Session transcripts, cron job definitions, and plugin-specific data do not transfer. Those are OpenClaw-specific and have different formats in Hermes.

Quick Migration

# Preview what would happen (no files changed)
hermes claw migrate --dry-run

# Run the full migration (includes API keys)
hermes claw migrate

# Exclude API keys (safer for shared machines)
hermes claw migrate --preset user-data

The migration reads from ~/.openclaw/ by default. If you have legacy ~/.clawdbot/ or ~/.moldbot/ directories, those are detected automatically.

Migration Options

Option	What It Does	Default
`--dry-run`	Preview without writing anything	off
`--preset full`	Include API keys and secrets	yes
`--preset user-data`	Exclude API keys	no
`--overwrite`	Overwrite existing Hermes files on conflicts	skip
`--migrate-secrets`	Include API keys explicitly	on with `--preset full`
`--source <path>`	Custom OpenClaw directory	`~/.openclaw/`
`--workspace-target <path>`	Where to place `AGENTS.md`	current directory
`--skill-conflict <mode>`	`skip`, `overwrite`, or `rename`	`skip`
`--yes`	Skip confirmation prompt	off

Step-by-Step Walkthrough

1. Dry Run First

Always preview before committing:

hermes claw migrate --dry-run

This shows you exactly what files would be created, overwritten, or skipped. Review the output carefully.

2. Run the Migration

hermes claw migrate

The tool will:

Detect your OpenClaw installation
Map config keys to Hermes equivalents
Merge memory files (deduplicating entries)
Copy skills to ~/.hermes/skills/openclaw-imports/
Migrate API keys (if --preset full)
Report what was done

3. Handle Conflicts

If a skill already exists in Hermes with the same name:

--skill-conflict skip (default): Leaves the Hermes version, skips the import
--skill-conflict overwrite: Replaces the Hermes version with the OpenClaw version
**--skill-conflict rename**: Creates a -imported` copy alongside the Hermes version

# Example: rename on conflict so you can compare
hermes claw migrate --skill-conflict rename

4. Verify After Migration

# Check your personality loaded
cat ~/.hermes/SOUL.md

# Check memory entries merged
cat ~/.hermes/memories/MEMORY.md | head -50

# Check skills imported
ls ~/.hermes/skills/openclaw-imports/

# Test the agent
hermes chat -q "What do you remember about me?"

What Doesn't Transfer

Item	Why	What to Do
Session transcripts	Different format	Archive manually if needed
Cron job definitions	Different scheduler	Recreate with `hermes cron`
Plugin configs	Plugin system changed	Reconfigure in Hermes
OpenClaw-specific features	May not exist yet	Check Hermes docs for equivalents

Config Key Mapping

For reference, here's how OpenClaw config maps to Hermes:

OpenClaw Config	Hermes Config	Notes
`agents.defaults.model`	`model`	String or `{primary, fallbacks}`
`agents.defaults.timeoutSeconds`	`agent.max_turns`	Divided by 10, capped at 200
`agents.defaults.verboseDefault`	`agent.verbose`	off / on / full
`agents.defaults.thinkingDefault`	`reasoning.mode`	off / low / high
`models.providers.*.baseUrl`	`custom_providers.*.base_url`	Direct mapping
`models.providers.*.apiType`	`custom_providers.*.api_type`	openai → chat_completions, anthropic → anthropic_messages

Troubleshooting

"No OpenClaw installation found"

Make sure your OpenClaw data is at ~/.openclaw/. If it's elsewhere:

hermes claw migrate --source /path/to/your/openclaw

Memory entries look duplicated

The migration deduplicates by content similarity, but if your OpenClaw memory had near-duplicates, they might not merge perfectly. Clean up manually:

# Edit memory directly
nano ~/.hermes/memories/MEMORY.md

Skills have import errors

OpenClaw skills may reference modules or patterns that don't exist in Hermes. Open the skill file and check the imports:

cat ~/.hermes/skills/openclaw-imports/skill-name/SKILL.md

Most skills work as-is since they're markdown-based instructions. Skills with code that imports OpenClaw-specific modules need manual updating.

What's Next

Want smarter memory? → Part 3: LightRAG Setup
Need mobile access? → Part 4: Telegram Setup
Want the agent to self-improve? → Part 5: On-the-Fly Skills

Part 3: LightRAG — Graph RAG That Actually Works

The Problem With Basic Memory

Hermes ships with vector-based memory search. It finds documents that are textually similar to your query. That works for simple lookups, but it has a fundamental ceiling: it finds what's similar, not what's connected.

Ask "what hardware decisions were made and why?" and vector search returns files that all mention GPUs. It can't traverse from a decision → the person who made it → the project it affected → the lesson learned afterward.

Graph RAG fixes this. It builds a knowledge graph (entities + relationships) alongside your vector database, then searches both simultaneously.

Naive RAG vs Graph RAG

	Naive RAG (Default)	Graph RAG (LightRAG)
Indexes	Text chunks as vectors	Entities, relationships, AND text chunks
Retrieves	Similar text (cosine similarity)	Connected knowledge (graph traversal + similarity)
Answers	"Here's what the docs say about X"	"Here's how X relates to Y, who decided Z, and why"
Scales	Degrades at 500+ docs (too many partial matches)	Improves with more docs (richer graph)
Cost	Cheap (embedding only)	More expensive upfront (LLM extracts entities) but cheaper at query time

LightRAG: The Best Graph RAG For Personal Use

LightRAG is an open-source graph RAG framework from HKU (EMNLP 2025 paper). It competes with Microsoft's GraphRAG at a fraction of the cost.

Why LightRAG over alternatives:

Tool	Graph	Vector	Web UI	Self-Hosted	API	Cost
LightRAG	Yes	Yes	Yes	Yes	REST API	Free
Microsoft GraphRAG	Yes	Yes	No	Yes	No	10-50x more
Graphiti + Neo4j	Yes	No (separate)	No (Neo4j browser)	Yes	Build your own	Free but manual
Plain vector search	No	Yes	No	Yes	Yes	Free

LightRAG does vector DB + knowledge graph in parallel during ingestion. One system, both capabilities.

Installation

Prerequisites

Python 3.11+
An LLM API key (for entity extraction during ingestion — OpenAI, Anthropic, or any OpenAI-compatible provider)
An embedding API key (Fireworks recommended for high-quality 4096-dim embeddings, or use local Ollama)

Install LightRAG

# Create a dedicated directory
mkdir -p ~/.hermes/lightrag
cd ~/.hermes/lightrag

# Clone LightRAG
git clone https://github.com/HKUDS/LightRAG.git
cd LightRAG

# Install dependencies
pip install -e ".[api]"

Set Up Environment

Create ~/.hermes/lightrag/.env:

# LLM for entity extraction (during ingestion)
LLM_BINDING=openai
LLM_MODEL=kimi-2.5                    # What we actually use — great quality/cost ratio
LLM_BINDING_API_KEY=<your-api-key>

# Embedding model (for vector storage)
EMBEDDING_BINDING=fireworks
EMBEDDING_MODEL=accounts/fireworks/models/qwen3-embedding-8b
EMBEDDING_API_KEY=<your-fireworks-api-key>

# Or use local Ollama (free, no API key needed):
# EMBEDDING_BINDING=ollama
# EMBEDDING_MODEL=nomic-embed-text

Security tip: Set restrictive permissions on this file: chmod 600 ~/.hermes/lightrag/.env

Entity Extraction Model — What to Use

This is the LLM that reads your documents and pulls out entities and relationships during ingestion. Quality here directly determines how good your knowledge graph is.

Model	Speed	Quality	Cost	Recommendation
Kimi 2.5	Fast	Excellent	Cheap	What we use. Great balance of quality, speed, and cost for entity extraction
Cerebras + Qwen 3	Blazing fast	Very good	Very cheap	Fastest option in the world. Cerebras inference at 2000+ tok/s makes bulk ingestion fly
GPT-4.1-mini	Fast	Good	Cheap	Solid fallback, well-tested
Claude Sonnet 4	Medium	Excellent	Mid-range	Overkill for ingestion but works great
Ollama local	Depends on GPU	Unpredictable	Free	Untested for this use case — might mess up entity extraction quality. Use at your own risk

Our recommendation: Kimi 2.5 for quality, Cerebras + Qwen 3 if you're ingesting a lot of documents and speed matters. Both are cheap and reliable. We haven't tested local Ollama for entity extraction — it's free but the extraction quality is unverified and you might get a messy graph.

Embedding quality matters. If you have a GPU with 8GB+ VRAM, run nomic-embed-text locally via Ollama for free. If you want the best quality, use Fireworks' Qwen3-Embedding-8B (4096 dimensions) — the search accuracy difference is dramatic.

Running the Server

Start the REST API

cd ~/.hermes/lightrag/LightRAG

# Start the API server (binds to localhost by default)
lightrag-server --host 127.0.0.1 --port 9623

The server starts on http://localhost:9623 with:

REST API for ingestion and querying
Web UI at http://localhost:9623/webui for browsing the knowledge graph
Health check at http://localhost:9623/health

Security warning: The LightRAG REST API has no built-in authentication. Always bind to 127.0.0.1 (localhost only) — never 0.0.0.0. If you need remote access, put it behind a reverse proxy (nginx, Caddy) with authentication, or use SSH tunneling. Anyone who can reach this port can query, ingest, or delete your knowledge graph data.

Run as a Background Service

# Using nohup
nohup lightrag-server --port 9623 > ~/.hermes/lightrag/server.log 2>&1 &

# Or use hermes to manage it
hermes background "cd ~/.hermes/lightrag/LightRAG && lightrag-server --port 9623"

Ingesting Your Knowledge

How Ingestion Works

Document (markdown, text, PDF, etc.)
    ↓
Chunking (text split into segments)
    ↓
┌─────────────────┐    ┌──────────────────┐
│ Embedding Model │    │ LLM Entity       │
│ (vector storage)│    │ Extraction       │
└────────┬────────┘    └────────┬─────────┘
         ↓                      ↓
   Vector Database       Knowledge Graph
   (similarity search)   (entity relationships)

For each document, LightRAG:

Chunks the text and embeds it (standard vector RAG)
Uses an LLM to extract entities (people, tools, projects, concepts) and relationships (who decided what, what depends on what)
Stores both in parallel — vectors for similarity, graph for structure

Ingest Documents via API

# Ingest a single file
curl -X POST http://localhost:9623/documents/upload \
  -F "file=@/path/to/your/document.md"

# Ingest a text string directly
curl -X POST http://localhost:9623/documents/text \
  -H "Content-Type: application/json" \
  -d '{"text": "Your knowledge content here...", "description": "Source description"}'

# Ingest all files in a directory
for file in ~/.hermes/memories/*.md; do
  curl -X POST http://localhost:9623/documents/upload -F "file=@$file"
  echo "Ingested: $file"
done

What to Ingest

Feed LightRAG everything your agent needs to "know":

Memory files — ~/.hermes/memories/*.md
Project docs — README files, design docs, decision logs
Chat summaries — Exported conversation summaries
Notes — Any markdown/text knowledge you want searchable
Code comments — Extracted from important codebases

Start with your memory files and project docs. These give the graph the most value — decisions, people, projects, and their relationships.

Querying the Graph

Query Modes

LightRAG has four query modes:

Mode	Best For	How It Works
`naive`	Simple keyword lookups	Vector search only (like basic RAG)
`local`	Specific entity facts	Entity-focused graph traversal
`global`	Cross-document relationships	Relationship-focused traversal
`hybrid`	General questions (default)	Both local + global combined

Query via API

# Hybrid query (recommended default)
curl -X POST http://localhost:9623/query \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What infrastructure decisions were made and why?",
    "mode": "hybrid",
    "only_need_context": false
  }'

# Local mode — specific entity facts
curl -X POST http://localhost:9623/query \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Tell me about the 5090 PC setup",
    "mode": "local"
  }'

# Global mode — relationship discovery
curl -X POST http://localhost:9623/query \
  -H "Content-Type: application/json" \
  -d '{
    "query": "How do the different projects relate to each other?",
    "mode": "global"
  }'

Get Just the Context (for your own LLM)

curl -X POST http://localhost:9623/query \
  -H "Content-Type: application/json" \
  -d '{
    "query": "What models are running on what hardware?",
    "mode": "hybrid",
    "only_need_context": true
  }'

This returns the raw context chunks without generating an answer — useful for feeding into your own pipeline or Hermes' LLM.

Integrating With Hermes

Create a LightRAG Skill

Create ~/.hermes/skills/research/lightrag/SKILL.md:

---
name: lightrag
description: Query the LightRAG knowledge graph for past decisions, infrastructure, projects, and lessons learned. Use before saying "I don't remember."
---

# LightRAG Knowledge Graph

Query the LightRAG knowledge graph for past decisions, infrastructure, projects, and lessons learned.

## When To Use
- User asks about past work, decisions, or "what happened with X"
- Need context on projects, hardware, or configurations
- Remembering lessons learned or past issues
- Any question where you'd say "I don't remember" — use this FIRST

## Usage
```bash
curl -s -X POST http://localhost:9623/query \
  -H "Content-Type: application/json" \
  -d '{"query": "YOUR QUERY", "mode": "hybrid", "only_need_context": true}'

Search Modes

hybrid (default): Combined vector + graph search
local: Entity-focused (specific facts)
global: Relationship-focused (how things connect)
naive: Vector-only (simple lookups)

Important

ALWAYS search this before saying "I don't remember"
Results supersede general knowledge about the setup
Reference entity names when citing results


### Query from a Script

Create `~/.hermes/skills/research/lightrag/scripts/lightrag_search.py`:

```python
#!/usr/bin/env python3
"""LightRAG search script for Hermes skill integration."""
import json
import sys
import urllib.request

def search(query: str, mode: str = "hybrid") -> str:
    url = "http://localhost:9623/query"
    payload = json.dumps({
        "query": query,
        "mode": mode,
        "only_need_context": True
    }).encode()
    
    req = urllib.request.Request(url, data=payload, headers={"Content-Type": "application/json"})
    try:
        with urllib.request.urlopen(req, timeout=30) as resp:
            result = json.loads(resp.read())
            return result.get("response", result.get("data", str(result)))
    except Exception as e:
        return f"LightRAG query failed: {e}"

if __name__ == "__main__":
    query = " ".join(sys.argv[1:]) if len(sys.argv) > 1 else ""
    if not query:
        print("Usage: lightrag_search.py <query>")
        sys.exit(1)
    print(search(query))

Optimizing Search Quality

1. Tune Entity Extraction

The quality of your graph depends on entity extraction. In LightRAG's config:

# More entities = richer graph, slower ingestion
entity_extract_max_gleaning: 5    # Default: 3. Higher = more thorough

# Chunk size affects entity density
chunk_token_size: 1200             # Default: 1200. Smaller = more entities per doc
chunk_overlap_token_size: 100      # Default: 100

2. Use High-Quality Embeddings

Embedding quality directly impacts vector search accuracy:

Model	Dimensions	Quality	Cost
nomic-embed-text (Ollama)	768	Good	Free (local)
Qwen3-Embedding-8B (Fireworks)	4096	Excellent	~$0.001/1K tokens
text-embedding-3-large (OpenAI)	3072	Very Good	~$0.00013/1K tokens

If search quality matters, use 4096-dimension embeddings. The difference between 768 and 4096 dims is like the difference between 720p and 4K — you catch details you'd otherwise miss.

3. Reindex After Bulk Changes

After ingesting a large batch of new documents:

# Check entity count
curl http://localhost:9623/graph/label/list | python3 -c "import sys,json; d=json.load(sys.stdin); print(f'{len(d)} entities')"

4. Use the Right Query Mode

Don't always default to hybrid. Use:

local when asking about a specific thing ("Tell me about the GPU setup")
global when asking about connections ("How do the projects relate?")
hybrid for general questions ("What decisions were made last week?")

5. Monitor and Prune

The Web UI at http://localhost:9623/webui lets you:

Browse the knowledge graph visually
See entity relationships
Identify orphaned or redundant entities

Web UI

Once the server is running, open http://localhost:9623/webui in your browser. You can:

Search the graph with any query mode
Visualize entity relationships as a network graph
Browse all entities and their connections
Inspect raw chunks and their source documents

Here's what a populated LightRAG knowledge graph looks like in the Web UI (screenshot from the LightRAG project):

The Web UI showing entity extraction, graph relationships, and document indexing. Once you ingest your own data, your graph fills up with your specific entities — people, projects, decisions, hardware, configs — all connected by the relationships LightRAG extracts automatically.

Troubleshooting

"Connection refused" on query

The server isn't running. Start it:

cd ~/.hermes/lightrag/LightRAG && lightrag-server --port 9623

Slow ingestion

Entity extraction is LLM-bound. Speed it up:

Use a faster model for ingestion (Cerebras + Qwen 3 is the fastest option, or Kimi 2.5)
Process documents in parallel batches
Use a local model if you have GPU capacity

Empty or irrelevant results

Check that documents were actually ingested (Web UI → entities)
Try different query modes (local vs global vs hybrid)
Rephrase your query — be more specific about entities
Check embedding model is actually running (curl http://localhost:11434/api/tags for Ollama)

Duplicate entities after re-ingestion

LightRAG merges similar entities automatically, but exact duplicates can happen. Use the Web UI to manually clean up, or reindex from scratch:

# Nuclear option: wipe and reingest
rm -rf ~/.hermes/lightrag/LightRAG/rag_storage/*
# Then re-ingest your documents

What's Next

Need mobile access? → Part 4: Telegram Setup
Want the agent to self-improve? → Part 5: On-the-Fly Skills

Part 4: Telegram Setup (Chat From Anywhere)

Why Telegram

Your agent is only useful if you can access it. Sitting at a terminal works until you need to:

Check something from your phone while away from your desk
Get notified when a long-running task finishes
Use Hermes in a group chat with your team
Send voice memos that get auto-transcribed and processed
Receive scheduled task results (cron jobs) on mobile

Telegram is the best messaging platform for Hermes bots — it supports text, voice, images, files, inline buttons, and group chats with minimal setup.

Step 1: Create a Bot via BotFather

Every Telegram bot requires an API token from @BotFather, Telegram's official bot management tool.

Open Telegram and search for @BotFather, or visit t.me/BotFather
Send /newbot
Choose a display name (e.g., "Hermes Agent") — this can be anything
Choose a username — this must be unique and end in bot (e.g., my_hermes_bot)
BotFather replies with your API token. It looks like this:

123456789:ABCdefGHIjklMNOpqrSTUvwxYZ

Keep your bot token secret. Anyone with this token can control your bot. If it leaks, revoke it immediately via /revoke in BotFather.

Step 2: Customize Your Bot (Optional)

These BotFather commands improve the user experience:

Command	Purpose
`/setdescription`	The "What can this bot do?" text shown before chatting
`/setabouttext`	Short text on the bot's profile page
`/setuserpic`	Upload an avatar for your bot
`/setcommands`	Define the command menu (the `/` button in chat)

For /setcommands, a useful starting set:

help - Show help information
new - Start a new conversation
sethome - Set this chat as the home channel
status - Show agent status

Step 3: Privacy Mode (Critical for Groups)

Telegram bots have privacy mode enabled by default. This is the single most common source of confusion.

With privacy mode ON, your bot can only see:

Messages that start with a / command
Replies directly to the bot's own messages
Service messages (member joins/leaves, pinned messages)

With privacy mode OFF, the bot receives every message in the group.

How to Disable Privacy Mode

Message @BotFather
Send /mybots
Select your bot
Go to Bot Settings → Group Privacy → Turn off

You must remove and re-add the bot to any group after changing the privacy setting. Telegram caches the privacy state when a bot joins a group — it won't update until removed and re-added.

Alternative: Promote the bot to group admin. Admin bots always receive all messages regardless of privacy settings.

Step 4: Find Your User ID

Hermes uses numeric Telegram user IDs to control access. Your user ID is not your username — it's a number like 123456789.

Method 1 (recommended): Message @userinfobot — it instantly replies with your user ID.

Method 2: Message @get_id_bot — another reliable option.

Save this number; you'll need it for the next step.

Step 5: Configure Hermes

Option A: Interactive Setup (Recommended)

hermes gateway setup

Select Telegram when prompted. The wizard asks for your bot token and allowed user IDs, then writes the configuration for you.

Option B: Manual Configuration

Add the following to ~/.hermes/.env:

TELEGRAM_BOT_TOKEN=<your-bot-token-from-botfather>
TELEGRAM_ALLOWED_USERS=<your-numeric-user-id>    # Comma-separated for multiple users

For groups, also add the group chat ID (negative number, like -1001234567890):

TELEGRAM_ALLOWED_CHATS=-1001234567890

Step 6: Start the Gateway

hermes gateway

The bot should come online within seconds. Send it a message on Telegram to verify.

Gateway Management

# Check gateway status
hermes gateway status

# Stop the gateway
hermes gateway stop

# Restart after config changes
hermes gateway restart

# Run as a system service (auto-start on boot)
hermes gateway install   # Sets up systemd/launchd service

Features Available on Telegram

Text Chat

Full conversation support — the bot processes your messages the same as the CLI.

Voice Messages

Send a voice memo and Hermes:

Auto-transcribes it using Whisper
Processes the transcription as a text message
Responds with text (or voice via TTS)

Image Analysis

Send a photo and Hermes analyzes it using vision models. Describe what you want to know about the image in the caption.

File Attachments

Send documents, code files, or data files — Hermes can read and process them.

Inline Buttons

For dangerous commands, Hermes shows confirmation buttons instead of executing immediately.

Slash Commands

The bot supports Telegram's native command menu (the / button in chat).

Scheduled Messages

Cron job results are delivered directly to your Telegram chat:

# Deliver cron results to Telegram
hermes cron create --deliver telegram "Check server status every hour" --schedule "every 1h"

Webhook Mode (For Cloud Deployments)

By default, Hermes uses long polling — the gateway makes outbound requests to Telegram. This works for local and always-on servers.

For cloud deployments (Fly.io, Railway, Render), webhook mode is better. These platforms auto-wake on inbound HTTP traffic but not on outbound connections.

Configuration

Add to ~/.hermes/.env:

TELEGRAM_WEBHOOK_URL=https://your-app.fly.dev
TELEGRAM_WEBHOOK_SECRET=<generate-with-command-below>

Generate a strong secret — never use a guessable value:

openssl rand -hex 32

Copy the output and paste it as your TELEGRAM_WEBHOOK_SECRET value.

Warning: A weak or default webhook secret lets attackers forge Telegram webhook requests and inject messages into your agent. Always use a cryptographically random value.

	Polling (default)	Webhook
Direction	Gateway → Telegram	Telegram → Gateway
Best for	Local, always-on servers	Cloud platforms
Extra config	None	`TELEGRAM_WEBHOOK_URL`
Idle cost	Machine must stay on	Machine can sleep

Multi-User Setup

To allow multiple users to interact with the bot:

TELEGRAM_ALLOWED_USERS=123456789,987654321,555555555

Each user gets their own conversation session. The bot tracks sessions per user ID.

Troubleshooting

Bot not responding

Check the token is correct: echo $TELEGRAM_BOT_TOKEN
Verify the gateway is running: hermes gateway status
Check logs: hermes gateway logs

Bot in group but not seeing messages

Privacy mode is still on. You must:

Disable privacy in BotFather (/mybots → Bot Settings → Group Privacy → Turn off)
Remove the bot from the group
Re-add the bot to the group

Voice messages not transcribed

Hermes needs ffmpeg for audio conversion. The installer includes it, but if you installed manually:

sudo apt install ffmpeg   # Ubuntu/Debian
brew install ffmpeg        # macOS

Rate limiting

Telegram limits bots to 30 messages/second to different chats and 20 messages/minute to the same group. If you're hitting limits, add a delay:

hermes config set telegram.rate_limit_delay 1

What's Next

Want the agent to self-improve? → Part 5: On-the-Fly Skills

Part 5: On-the-Fly Skills (Let Hermes Build Its Own Playbook)

What Are Skills

Skills are procedural knowledge — step-by-step instructions that teach Hermes how to handle specific tasks. Unlike memory (which is factual), skills are how-to guides the agent follows automatically.

Skills vs Memory:

	Skills	Memory
What	How to do things	What things are
When	Loaded on demand, only when relevant	Injected every session automatically
Size	Can be large (hundreds of lines)	Should be compact (key facts only)
Cost	Zero tokens until loaded	Small but constant token cost
Examples	"How to deploy to Kubernetes"	"User prefers dark mode, lives in EST"
Who creates	You, the agent, or installed from Hub	The agent, based on conversations

Rule of thumb: If you'd put it in a reference document, it's a skill. If you'd put it on a sticky note, it's memory.

The Skill Creation Workflow

Hermes can create skills itself. Here's how it works:

1. Do a Complex Task

Ask Hermes to do something multi-step. For example:

Set up a monitoring script that checks my server health every 5 minutes
and alerts me on Telegram if CPU goes above 90% or memory above 80%.

Hermes will:

Research the best approach
Write the script
Test it
Set up the cron job
Fix any issues along the way

2. Hermes Offers to Save It

After completing a complex task (5+ tool calls), fixing a tricky error, or discovering a non-trivial workflow, Hermes will offer:

This was a multi-step process. Want me to save this as a skill
so I can reuse it next time?

3. Say Yes

The agent uses skill_manage to create a new skill file at ~/.hermes/skills/<category>/<skill-name>/SKILL.md. This file contains:

When to use — the trigger conditions
Exact steps — commands, files, configurations
Pitfalls — problems encountered and how to fix them
Verification — how to confirm it worked

4. It's Available Immediately

The skill appears in skills_list and becomes available as a slash command. Next time you (or the agent) encounter a similar task, the skill is loaded automatically.

How to Ask Hermes to Create a Skill

Direct Request

Just ask:

Create a skill for deploying Docker containers to my server.
Include the build, push, SSH deploy, and health check steps.

Hermes will:

Research the best deployment workflow
Create the skill directory at ~/.hermes/skills/
Write SKILL.md with the full procedure
Add reference files, templates, or scripts if needed
Test that it works

After Solving a Problem

If Hermes just solved a tricky problem for you:

Save that as a skill so you remember how to do it next time.

The agent captures:

The exact steps taken
The errors encountered and fixes
The configuration needed
Edge cases discovered

Iterative Improvement

If a skill is outdated or incomplete:

That skill doesn't cover the new deployment method. Update it
with what we just learned.

Hermes patches the skill with new information using skill_manage(action='patch').

Skill Structure

Every skill is a directory with a SKILL.md file:

~/.hermes/skills/
├── my-category/
│   ├── my-skill/
│   │   ├── SKILL.md              # Main instructions (required)
│   │   ├── references/           # Supporting docs (optional)
│   │   │   ├── api-docs.md
│   │   │   └── examples.md
│   │   ├── templates/            # Template files (optional)
│   │   │   └── config.yaml
│   │   └── scripts/              # Executable scripts (optional)
│   │       └── setup.sh
│   └── another-skill/
│       └── SKILL.md
└── openclaw-imports/             # Migrated from OpenClaw
    └── old-skill/
        └── SKILL.md

SKILL.md Format

---
name: my-skill
description: Brief description of what this skill does
version: 1.0.0
metadata:
  hermes:
    tags: [deployment, docker, devops]
    category: my-category
---

# My Skill

## When to Use
Use this skill when the user asks to deploy containers or manage Docker services.

## Procedure
1. Check Docker is running: `docker ps`
2. Build the image: `docker build -t app:latest .`
3. Push to registry: `docker push registry/app:latest`
4. SSH to server and pull: `ssh server 'docker pull registry/app:latest && docker-compose up -d'`
5. Health check: `curl -f http://server:8080/health`

## Pitfalls
- Docker build fails if Dockerfile has COPY paths wrong — fix by checking working directory
- SSH needs key-based auth — set up with `ssh-keygen` and `ssh-copy-id`
- Health check may take 10s to respond — add retry logic

## Verification
Run `docker ps` on the server and confirm the container is `Up` and healthy.

Using Skills

Via Slash Command

Every skill becomes a slash command automatically:

/my-skill deploy the latest version to production

Via Natural Conversation

Just ask Hermes to use a skill:

Use the docker-deploy skill to push the new build.

Hermes loads the skill via skill_view and follows its instructions.

Automatic Loading

Hermes scans available skills at session start. When your request matches a skill's "When to Use" conditions, it loads automatically — you don't need to explicitly invoke it.

Managing Skills

List All Skills

/skills
# Or
hermes skills list

Search for a Skill

/skills search docker
/skills search deployment

View a Skill's Content

/skills view my-skill

Enable/Disable Per Platform

hermes skills

This opens an interactive TUI where you can enable or disable skills per platform (CLI, Telegram, Discord, etc.). Useful when you want certain skills only available in specific contexts.

Install from the Hub

Official optional skills (heavier or niche):

/skills install official/research/arxiv
/skills install official/creative/songwriting-and-ai-music

Update a Skill

If a skill is outdated or missing steps:

Update the docker-deploy skill — we learned that the health check
needs a 30-second timeout, not 10.

Hermes patches the skill with skill_manage(action='patch').

Real-World Skill Examples

Example 1: Server Monitoring

Create a skill that monitors my server: check CPU, memory, and disk
usage via SSH, log results to a CSV, and alert on Telegram if anything
exceeds thresholds.

Hermes creates a skill with:

SSH connection commands
Resource check scripts
CSV logging format
Telegram alert integration
Threshold configuration

Example 2: Code Review

Create a skill for reviewing Python pull requests. It should check
for security issues, performance problems, and style violations.

Hermes creates a skill with:

git diff analysis steps
Security pattern checks
Performance anti-pattern detection
Style guide references

Example 3: Lead Research

Create a skill that researches companies: find their website, check
LinkedIn for key contacts, look at recent news, and compile a one-page summary.

Hermes creates a skill with:

Web search queries to use
LinkedIn search patterns
News aggregation approach
Summary template

Tips for Better Skills

Be specific about the task. "Deploy Docker containers" is too vague. "Deploy a Python Flask app to a VPS using Docker Compose with health checks" gives the agent enough detail to write a precise skill.

Include examples. When asking for a skill, show an example of the desired output. This helps the agent write better templates.

Let the agent discover pitfalls. Don't prescribe the exact steps. Let Hermes figure out the workflow and capture what goes wrong — those pitfall notes are the most valuable part of the skill.

Update skills when they go stale. If you use a skill and hit issues not covered by it, tell Hermes to update it with what you learned. Skills that aren't maintained become liabilities.

Use categories. Organize skills into subdirectories (~/.hermes/skills/devops/, ~/.hermes/skills/research/, etc.). This keeps the list manageable and helps the agent find relevant skills faster.

Keep skills focused. A skill that tries to cover "all of DevOps" will be too long and too vague. A skill that covers "deploy a Python app to Fly.io" is specific enough to be genuinely useful.

How Hermes Decides to Save Skills

The agent saves skills automatically after:

Complex tasks (5+ tool calls) — multi-step workflows worth preserving
Tricky error fixes — debugging steps that took iteration to solve
Non-trivial discoveries — new approaches or configurations found during work
User request — when you explicitly say "save this as a skill"

The agent uses skill_manage(action='create') to write the skill, including:

Trigger conditions
Numbered steps with exact commands
Pitfalls section (from actual errors encountered)
Verification steps

What's Next

You've now got the full picture:

Part 1: Setup — Install and configure
Part 2: OpenClaw Migration — Bring your old data
Part 3: LightRAG — Graph-based knowledge
Part 4: Telegram — Mobile access
Part 5: On-the-Fly Skills — Self-improving workflows

Start with setup, add what you need, and let Hermes build the rest.

Note: Based on the official Hermes Agent documentation and real production usage. No private credentials, API keys, or personal data included.

Name		Name	Last commit message	Last commit date
Latest commit History 26 Commits
.github		.github
benchmarks		benchmarks
diagrams		diagrams
docs		docs
screenshots		screenshots
scripts		scripts
skills		skills
templates		templates
CHANGELOG.md		CHANGELOG.md
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
CONTRIBUTING.md		CONTRIBUTING.md
ECOSYSTEM.md		ECOSYSTEM.md
LICENSE		LICENSE
README-ja.md		README-ja.md
README-zh.md		README-zh.md
README.md		README.md
ROADMAP.md		ROADMAP.md
part1-setup.md		part1-setup.md
part10-soul-antipatterns.md		part10-soul-antipatterns.md
part11-gateway-recovery.md		part11-gateway-recovery.md
part12-web-dashboard.md		part12-web-dashboard.md
part13-tool-gateway.md		part13-tool-gateway.md
part14-fast-mode-watchers.md		part14-fast-mode-watchers.md
part15-new-platforms.md		part15-new-platforms.md
part16-backup-debug.md		part16-backup-debug.md
part17-mcp-servers.md		part17-mcp-servers.md
part18-coding-agents.md		part18-coding-agents.md
part19-security-playbook.md		part19-security-playbook.md
part2-openclaw-migration.md		part2-openclaw-migration.md
part20-observability.md		part20-observability.md
part21-remote-sandboxes.md		part21-remote-sandboxes.md
part3-lightrag-setup.md		part3-lightrag-setup.md
part4-telegram-setup.md		part4-telegram-setup.md
part5-creating-skills.md		part5-creating-skills.md
part6-context-compression.md		part6-context-compression.md
part7-memory-system.md		part7-memory-system.md
part8-subagent-patterns.md		part8-subagent-patterns.md
part9-custom-models.md		part9-custom-models.md

Folders and files

Latest commit

History

Repository files navigation

Hermes Optimization Guide

The End-to-End Hermes Guide — docs + runnable artifacts

Install Everything (one command)

Repo Map

Architecture at a glance

Pick Your Path

🎯 "I just want it working in 10 minutes"

📱 "I want a Telegram bot that's actually useful"

🤖 "I want to drive Claude Code / Codex / Gemini from my phone"

💼 "I'm running this in production"

🧠 "I want the most capable agent possible, cost be damned"

💰 "I want the cheapest possible agent that still works"

🛡️ "I'm worried about prompt injection (you should be)"

What's New (April 10–17, 2026)

v0.9.0 — 2026.4.13 — "Everywhere"

v0.10.0 — 2026.4.16 — "Tool Gateway"

🔥 Cooking on main (past 72 hours, targeting v0.11)

🆕 Brand-new parts in this guide (April 17)

Viral model / provider developments (past 72h, now in the guide)

Table of Contents

The Problem

What This Fixes

Prerequisites

How the Pieces Fit Together

Quick Start

Part 1: Setup (Stop Fumbling With Installation)

The Install

Linux / macOS / WSL2

What the Installer Does

After Installation

First-Run Configuration

1. Choose Your LLM Provider

Local Models (Ollama)

2. Set Your API Keys

3. Configure Toolsets

Key Config Options

Model Settings

Agent Behavior

Context Management

SOUL.md — Give Your Agent a Personality

What Belongs in SOUL.md

The Molty Prompt

What Good Looks Like

Why This Works

File Locations

Verify Your Setup

Updating

What's Next

Part 2: OpenClaw Migration (Don't Leave Your Knowledge Behind)

Why Migrate

Quick Migration

Migration Options

Step-by-Step Walkthrough

1. Dry Run First

2. Run the Migration

3. Handle Conflicts

4. Verify After Migration

What Doesn't Transfer

Config Key Mapping

Troubleshooting

"No OpenClaw installation found"

Memory entries look duplicated

Skills have import errors

What's Next

Part 3: LightRAG — Graph RAG That Actually Works

The Problem With Basic Memory

Naive RAG vs Graph RAG

LightRAG: The Best Graph RAG For Personal Use

Installation

Prerequisites

Install LightRAG

Set Up Environment

Entity Extraction Model — What to Use

Running the Server

Start the REST API

Run as a Background Service

🔥 Cooking on `main` (past 72 hours, targeting v0.11)