HERMES AGENT + NOTEBOOKLM + OBSIDIAN:BUILD A 3-AGENT RESEARCH DEPARTMENT THAT GETS SMARTER EVERY DAY

One agent doing research, analysis, and briefing at the same time produces mediocre results. Context gets polluted. The agent confuses what to find with what to analyze with what to report. Priorities blur. Quality drops with every added responsibility.

Three separate agents, each doing one job, produce compounding results. Scout finds signals. Analyst synthesizes meaning. Briefer delivers action items. Each profile has its own SOUL.md, its own model, its own memory, its own skills. Isolated. Focused. Coordinated through a shared Obsidian vault.

This article shows how to build that system. All three profiles, their SOUL.md templates, their cron jobs, the NotebookLM connection, the Obsidian wiki, and the coordination pattern that ties them together.

Total cost: $19-27/month depending on model choice. Setup time: one evening for the standard configuration.

Who this is for:

→ Solo founders tracking competitors and market trends → Content creators who need daily research for their niche → Agency owners monitoring multiple industries for clients → Researchers following academic papers and industry developments → Startup teams building competitive intelligence without hiring an analyst

If you spend more than 30 minutes per day on manual research, reading newsletters, or checking competitor updates, this setup pays for itself in the first week.

All technical details verified against Hermes Agent v0.16.0 official documentation.

WHY THREE AGENTS, NOT ONE

A single Hermes profile handling research end-to-end carries every source, every analysis note, and every briefing draft in one context window. By day 3, the context is heavy with accumulated research that has nothing to do with this morning’s brief. By week 2, the agent has 40+ skills covering everything from arXiv parsing to Telegram formatting. Tool Search helps but the fundamental problem remains: one identity trying to be three different workers.

Profiles solve this at the architecture level. Each profile in Hermes is a fully isolated agent: own SOUL.md, own config.yaml, own memory, own skills, own cron jobs. They share nothing by default. What they share by design is a directory: the Obsidian vault where Scout deposits raw findings, Analyst writes synthesized notes, and Briefer reads each morning.

Three profiles. Three clear jobs. One shared knowledge base.

Fastest setup path: Hermes Desktop app.

The Desktop app (v0.16.0) has a built-in Profile Builder. No terminal needed:

hermes dashboard → Profiles → Build

Five-step wizard for each profile: Identity → Model → Skills → MCPs → Review. Create all three profiles in 15 minutes. The screenshots in this article follow the Desktop path.

You can also create profiles via CLI:

bashhermes profile create scout hermes profile create analyst hermes profile create briefer

Both paths produce the same result. Desktop is faster for first-time setup. CLI is faster once you know what you want.

YanXbt@IBuzovskyi·Jun 12 ArticleHERMES AGENT SOUL.MD: WHY 50 LINES MATTER MORE THAN YOUR MODEL (COMPLETE GUIDE)SOUL.md is the most important file in your Hermes Agent setup. It occupies slot #1 in the system prompt. Every turn, every session, every profile reads it first. It defines who the agent is before…855424158K

Each profile needs its own SOUL.md. The identity guide covers structure, token impact, and advanced templates.

THE THREE ROLES

SCOUT — finds signals

The Scout checks sources on a schedule and drops raw findings into an inbox. No analysis. No synthesis. No opinion. Raw signal only.

Profile setup:

bashhermes profile create scout

HERMES AGENT DESKTOP SCOUT PROFILE

HERMES AGENT DESKTOP SCOUT PROFILE

SOUL.md for Scout:

markdown# Soul You are a research scout. Your job is to find signals. You do not analyze. You do not summarize. You find relevant information and save it.

Voice

Terse. File names and one-line descriptions only. No commentary. No recommendations.

Operations

Search the sources listed in your cron jobs. For each finding: save the full text as a markdown file to ~/research/inbox/ with format: YYYY-MM-DD-source-keyword.md Include the source URL on the first line.

Restrictions

Never analyze or synthesize what you find. Never write more than 3 lines of your own text per file. Never delete files from the inbox. Never modify files written by other profiles.

Model: GPT-5.5 (cheap, high volume search tasks). Scout does low-reasoning work. Cheap model is the right choice.

For X/Twitter search specifically: GPT-5.5 cannot search X natively. Two options:

→ Use the xurl skill (X API integration, works with any model, requires X Developer App credentials) → Switch Scout to Grok via SuperGrok OAuth (native X search built into the model)

If X monitoring is a core part of your research, Grok as Scout model simplifies the setup. If you only need web + arXiv + RSS, GPT-5.5 handles everything.

Tools: web search, X search (xurl), RSS feeds, arXiv API.

Cron jobs:

/cron add “every 3h”
–prompt “Search X for posts about [your niche keywords] with more than 50 likes in the last 3 hours. Save each relevant finding as a markdown file to ~/research/inbox/. Include source URL.”
–deliver telegram

/cron add “every morning 7am”
–prompt “Check arXiv for new papers in [cs.AI, cs.CL] from the last 24 hours. Save titles, abstracts, and URLs to ~/research/inbox/.”
–deliver telegram

/cron add “every day 9am”
–prompt “Check these competitor URLs for changes: [url1, url2, url3]. If any page changed since last check, save the diff to ~/research/inbox/.”
–script competitor-diff.py

/cron add “every monday 8am”
–prompt “Scan Product Hunt for AI launches from the past 7 days. Save top 10 by upvotes to ~/research/inbox/.”
–deliver telegram

HERMES AGENT DESKTOP CRON JOBS SET

HERMES AGENT DESKTOP CRON JOBS SET

Most Scout crons use wakeAgent gates. The competitor diff script checks for changes before waking the agent. No changes = zero tokens.

ANALYST — synthesizes meaning

The Analyst processes raw findings from Scout, runs them through NotebookLM for deep synthesis, and writes structured notes to the Obsidian wiki. This is where raw signals become usable knowledge.

Profile setup:

bashhermes profile create analyst

HERMES AGENT DESKTOP ANLYST PROFILE SETUP

HERMES AGENT DESKTOP ANLYST PROFILE SETUP

SOUL.md for Analyst:

markdown# Soul You are a research analyst. Your job is to synthesize. You turn raw findings into structured knowledge. You verify claims. You flag contradictions. You connect ideas across sources.

Voice

Precise. Evidence-based. Every claim tagged with confidence level: [verified] [likely] [unverified] [conflicting]. Use tables for comparisons. Use bullet points for lists. Cite sources for every factual claim.

Operations

Process files from ~/research/inbox/. For each batch:

1. Feed sources to NotebookLM for cross-source synthesis (if NotebookLM unavailable, run synthesis directly via /goal)
2. Extract key insights from the synthesis
3. Write structured notes to ~/obsidian-wiki/ using the LLM Wiki skill
4. Tag each entry with confidence level
5. Flag contradictions with existing wiki entries
6. Move processed files to ~/research/processed/

Restrictions

Never present unverified claims as facts. Never skip the confidence tagging step. Never write to the wiki without source attribution. Never delete wiki entries. Update or flag only. Never modify files in ~/research/inbox/ that were not created by Scout.

Model: Claude Sonnet 4 (strong reasoning for synthesis). This is where quality matters. The Analyst writes the knowledge that Briefer reads every morning.

Tools: NotebookLM MCP, Obsidian skill (bundled), web search (for verification), file tools.

Cron job:

/cron add “every day 10am”
–script check-inbox.py
–prompt “Process all files in ~/research/inbox/. Feed them to NotebookLM for synthesis. Extract key insights. Write structured notes to Obsidian wiki. Tag confidence levels. Flag contradictions. Move processed files to ~/research/processed/.”
–deliver telegram

HERMES AGENT DESKTOP ANALYST CRON

HERMES AGENT DESKTOP ANALYST CRON

The inbox check script (wakeAgent gate):

Save as ~/.hermes/scripts/check-inbox.py:

python#!/usr/bin/env python3 import os, json

inbox = os.path.expanduser(“~/research/inbox”) files = [f for f in os.listdir(inbox) if f.endswith(‘.md’)] if os.path.exists(inbox) else []

if files: print(json.dumps({“wakeAgent”: True})) print(f“{len(files)} new files in inbox:“) for f in files: print(f” {f}“) else: print(json.dumps({“wakeAgent”: False}))

Empty inbox = zero tokens. New files = Analyst wakes and processes.

BRIEFER — delivers action items

The Briefer reads the Obsidian wiki every morning, cross-references with your current projects and calendar, and delivers a prioritized brief to Telegram.

Profile setup:

bashhermes profile create briefer

HERMES AGENT DESKTOP BRIEFER PROFILE SETUP

HERMES AGENT DESKTOP BRIEFER PROFILE SETUP

SOUL.md for Briefer:

markdown# Soul You are a briefing officer. Your job is to deliver a short, prioritized, actionable morning brief. You do not research. You do not analyze. You read what Analyst wrote and tell me what matters today.

Voice

5 bullets maximum. Each bullet: one finding, why it matters to me, suggested action. No preamble. No summary of the summary. Start with the most important item.

Operations

Every morning:

1. Read recent entries in ~/obsidian-wiki/ (last 24 hours)
2. Cross-reference with my current projects (check MEMORY.md and kanban board)
3. Prioritize by relevance to this week’s goals
4. Deliver 5-bullet brief to Telegram
5. End with total token spend this week

Restrictions

Never exceed 5 bullets in the brief. Never include items older than 48 hours unless flagged [urgent]. Never repeat items from yesterday’s brief unless status changed.

Model: GPT-5.5 (concise summaries, low token volume). Briefer does light synthesis and formatting. One brief per day. Cheap model handles this well.

Tools: Obsidian skill (read), session recall, file tools.

Cron job:

/cron add “every day 8am”
–prompt “Read the Obsidian wiki entries from the last 24 hours. Cross-reference with my current projects and this week’s goals. Deliver a 5-bullet prioritized brief. Most important item first. End with token spend this week.”
–deliver telegram

HERMES AGENT DESKTOP BRIEFER CRON

HERMES AGENT DESKTOP BRIEFER CRON

THE NOTEBOOKLM CONNECTION

NotebookLM is what gives the Analyst depth. Instead of the Analyst synthesizing sources through its own reasoning (good but limited by context window), NotebookLM ingests all sources, cross-references across them, and produces synthesis that draws from the full corpus.

What NotebookLM adds to the pipeline:

→ Multi-source synthesis (connects ideas across 50+ sources) → Audio overviews (podcast-style digest of your research) → Question answering from your curated source library → Zero hallucination from verified sources

How to connect:

The tool is notebooklm-mcp-cli by jacob-bd. 35 MCP tools for programmatic NotebookLM access: https://github.com/jacob-bd/notebooklm-mcp-cli

YanXbt@IBuzovskyi·May 28HERMES AGENT JUST GOT A RESEARCH BRAIN.

connect it to NotebookLM and it becomes a 247 research system that builds knowledge over time.

it resources. connects ideas across sources. uses that context in every real task.

setup in 4 steps:

1. install Hermes with MCP enabled 2.Show moreQuoteYanXbt@IBuzovskyi·May 28 ArticleHERMES AGENT: THE COMPLETE GUIDE. From Zero to Self-Improving AI EmployeeI’ve been running Hermes Agent 24/7 for months. This is everything I learned: installation, models, use cases, the dashboard 99% of people use wrong, self-improvement, security. All of it.

Bookmark…562111.6K182K

Install:

bashpip install notebooklm-mcp-cli nlm login

This opens a browser for Google OAuth. Log in with your Google account.

Setup via Desktop app (fastest):

hermes dashboard → MCP → Add Server

Name: notebooklm Transport: stdio Command: nlm Arguments: mcp serve

Save → Test Connection (should show 35 available tools)

Then assign to Analyst profile only:

hermes dashboard → Profiles → analyst → MCPs Enable notebooklm server for this profile

Or do it all at once through Profile Builder:

hermes dashboard → Profiles → Build Step 1: Identity (paste Analyst SOUL.md) Step 2: Model (Claude Sonnet) Step 3: Skills (enable llm-wiki) Step 4: MCPs → Add notebooklm server Step 5: Review → Deploy

One wizard. NotebookLM connected to Analyst profile.

Setup via CLI (alternative):

Generate config for Hermes:

bashnlm setup add json

Copy the generated JSON config into your Analyst profile’s MCP settings:

bashhermes mcp

add the notebooklm-mcp server

paste the config from nlm setup

What the Analyst can do through NotebookLM:

nlm notebook create “Weekly Research” nlm source add ~/research/inbox/file1.md nlm source add ~/research/inbox/file2.md nlm source add-research “query about your niche”

The Analyst runs these through the MCP tools automatically.

Honest caveat:

NotebookLM does not have a public API for the consumer product as of June 2026. The Enterprise product does. **notebooklm-mcp-cli **uses a Playwright-based browser automation wrapper under the hood. If Google changes an internal endpoint, the wrapper can break.

This is a real limitation. Plan for it.

The fallback path:

Add this to the Analyst SOUL.md:

If NotebookLM connection fails, run synthesis directly using /goal with this structure: “synthesize these [N] sources. find connections. flag contradictions. write to Obsidian wiki.”

Hermes /goal with Claude Sonnet produces strong synthesis on its own. NotebookLM makes it deeper. The fallback makes it reliable.

OBSIDIAN AS THE SHARED KNOWLEDGE BASE

Obsidian is the only component all three profiles touch. It is the shared memory layer.

Hermes ships with a bundled LLM Wiki skill based on Andrej Karpathy’s LLM Wiki pattern. It compiles knowledge into interlinked markdown files. Cross-references stay linked. Contradictions get flagged automatically.

Vault structure:

~/obsidian-wiki/ ├── inbox/ # raw findings from Scout (temporary) ├── sources/ # processed source pages ├── synthesis/ # Analyst’s structured notes ├── briefs/ # archived morning briefs ├── entities/ # people, companies, products ├── contradictions/ # flagged conflicts └── .last-pushed # timestamp for sync tracking

Setup:

Set the wiki path for each profile through the Dashboard:

hermes dashboard → Config → search “WIKI” WIKI_PATH = ~/obsidian-wiki OBSIDIAN_VAULT_PATH = ~/obsidian-wiki

Repeat for each profile (Scout, Analyst, Briefer).

First run: wiki initialization.

On first use, the LLM Wiki skill detects an empty directory and asks for a domain. This builds SCHEMA.md with tag taxonomy and conventions for your wiki. Example response:

AI agents, automation frameworks, and solo founder tooling.

focus areas:

• agent architecture and ecosystem
• competitor frameworks and comparisons
• AI model releases and benchmarks
• automation workflows and multi-agent systems
• token economics and cost optimization

Replace with your actual niche. The skill creates SCHEMA.md once and uses it for all future indexing. You can edit SCHEMA.md later if your focus shifts.

All three profiles point to the same directory. Scout writes to inbox/. Analyst reads inbox/, writes to sources/ and synthesis/. Briefer reads synthesis/.

What you see in Obsidian:

Open the vault in Obsidian on your laptop. Graph view shows nodes growing as the system runs. Backlinks connect related entries. Tags organize by confidence level. The knowledge graph builds itself overnight.

YanXbt@IBuzovskyi·Jun 15HERMES AGENT SHIPS WITH A BUNDLED SKILL FOR ANDREJ KARPATHY’S LLM WIKI PATTERN. A SELF-IMPROVING KNOWLEDGE BASE THAT GROWS EVERY TIME YOU FEED IT.

mentioned this briefly in the overnight workflow article. here is the full breakdown.

what it is: a self-improving knowledge baseShow moreQuoteYanXbt@IBuzovskyi·Jun 14 ArticleHERMES AGENT BUILDS ITSELF WHILE YOU SLEEP. THE COMPLETE GUIDE TO THE 9-HOUR OVERNIGHT WORKFLOW.Most AI agents wait for you to type. Hermes Agent does not. And it does something none of the others do: it gets smarter overnight. Between 11PM and 8AM, a properly configured Hermes setup monitors…132925029K

The LLM Wiki skill handles indexing, cross-referencing, and contradiction detection.

HOW THEY COORDINATE

No Kanban needed for this setup. File-based coordination with wakeAgent gates.

SCOUT (runs every 3 hours): → searches sources → drops markdown files to ~/research/inbox/ → notifies on Telegram what was found

ANALYST (runs daily at 10am): → wakeAgent script checks ~/research/inbox/ → empty inbox? sleep. zero tokens. → files found? wake up. process through NotebookLM. → write to Obsidian wiki → move processed files to ~/research/processed/

BRIEFER (runs daily at 8am): → reads recent Obsidian wiki entries → cross-references with projects and goals → delivers 5-bullet brief to Telegram

ALL 3 PROFILES TOGETHER IN THE HERMES DESKTOP APP

ALL 3 PROFILES TOGETHER IN THE HERMES DESKTOP APP

Why file-based and not Kanban:

Kanban is powerful but adds overhead for a pipeline this linear. Scout → Analyst → Briefer is a straight line. File inbox + wakeAgent gate is simpler, cheaper (zero dispatcher overhead), and easier to debug (just check the inbox folder).

If you later add more roles (Code Reviewer, Content Writer, Outreach Agent), Kanban becomes worth it. For three profiles in a pipeline, files are enough.

SETUP

Three tiers. Hermes handles most of the configuration. You tell it what you want.

**BASIC **

Scout + Briefer only. No Analyst. No NotebookLM. No Obsidian.

Step 1 — Create two profiles in Dashboard:

hermes dashboard → Profiles → Build

Create Scout: paste Scout SOUL.md → Model: GPT-5.5 (or Grok for X search) → Deploy. Create Briefer: paste Briefer SOUL.md → Model: GPT-5.5 → Deploy.

Step 2 — Tell each profile what to do:

Open Scout and say:

“Set up a cron job that runs every 3 hours. Search web for [your niche keywords]. Save findings as markdown files to ~/research/inbox/. Include source URL on the first line of each file. Deliver confirmation to Telegram.”

Open Briefer and say:

“Set up a cron job that runs every day at 8am. Read all files in ~/research/inbox/. Deliver a 5-bullet prioritized brief to Telegram. Most important item first.”

Hermes creates the cron jobs, folders, and configs.

Step 3 — Connect Telegram:

• Open Telegram, message @BotFather, send /newbot

• Name the bot, copy the token

• Message @userinfobot to get your user ID

hermes dashboard → Channels → Telegram

Paste bot token. Enter your user ID. Save. Restart gateway.

One bot handles all profiles. Scout, Analyst, and Briefer all deliver to the same chat.

STANDARD

All three profiles. Obsidian wiki. No NotebookLM.

Step 1 — Create three profiles in Dashboard:

hermes dashboard → Profiles → Build

Create Scout: paste Scout SOUL.md → Model: GPT-5.5 (or Grok) → enable xurl skill if monitoring X → Deploy. Create Analyst: paste Analyst SOUL.md → Model: Claude Sonnet → enable llm-wiki skill → Deploy. Create Briefer: paste Briefer SOUL.md → Model: GPT-5.5 → enable llm-wiki skill → Deploy.

Step 2 — Set wiki path for each profile:

hermes dashboard → Config → search “WIKI” WIKI_PATH = ~/obsidian-wiki OBSIDIAN_VAULT_PATH = ~/obsidian-wiki

Repeat for each profile. On first use the LLM Wiki skill asks for your domain (see Obsidian section above).

Step 3 — Tell each profile what to do:

Open Scout and say:

“Set up two cron jobs:

1. Every 3 hours: search web for [your niche keywords]. Save findings as markdown to ~/research/inbox/. Include source URL on the first line.

2. Every morning at 7am: check arXiv for new papers in [your topics]. Save titles, abstracts, and URLs to ~/research/inbox/.

Deliver confirmation to Telegram for both.“

Open Analyst and say:

“Set up a cron job that runs every day at 10am. Check ~/research/inbox/ for new markdown files. Use a wakeAgent script so you only fire when there are actually new files. If files found: synthesize them using /goal. Write structured notes to Obsidian wiki using the llm-wiki skill. Tag each entry with confidence level: verified, likely, unverified, or conflicting. Flag contradictions with existing wiki entries. Move processed files to ~/research/processed/. Deliver summary to Telegram.”

Open Briefer and say:

“Set up a cron job that runs every day at 8am. Read Obsidian wiki entries from the last 24 hours. Cross-reference with my current projects and goals. Deliver a 5-bullet prioritized brief to Telegram. Most important item first. End with token spend this week.”

Hermes creates the scripts, cron jobs, and folders for you.

Step 4 — Connect Telegram:

hermes dashboard → Channels → Telegram

@BotFather → /newbot → copy token → @userinfobot → copy user ID → paste both in Dashboard → restart gateway. One bot, all three profiles deliver to the same chat.

Tell Analyst: “drop a test file in the inbox and process it.” Verify the wiki entry appears in ~/obsidian-wiki/. Next morning: check Telegram for the Briefer output.

Step 5 — Test:

Tell Analyst: “drop a test file in the inbox and process it.” Verify the wiki entry appears in ~/obsidian-wiki/. Next morning: check Telegram for the Briefer output.

**ADVANCED **

All three profiles. Obsidian wiki. NotebookLM connected. Competitive analysis.

Step 1 — Create three profiles:

Same as Standard Step 1. Except when creating Analyst, skip MCPs for now (add NotebookLM in Step 3).

Step 2 — Set wiki path:

Same as Standard Step 2.

Step 3 — Connect NotebookLM to Analyst:

This is the one manual step. NotebookLM has no official API, so the connection uses a community MCP wrapper:

bashpip install notebooklm-mcp-cli nlm login

Browser opens for Google OAuth. Log in.

Then add to Analyst through Dashboard:

hermes dashboard → MCP → Add Server

Name: notebooklm Transport: stdio Command: nlm Arguments: mcp serve

Save → Test Connection (should show 35 tools)

Assign to Analyst profile only:

hermes dashboard → Profiles → analyst → MCPs Enable notebooklm server.

Step 4 — Tell each profile what to do:

Open Scout and say:

“Set up four cron jobs:

1. Every 3 hours: search web for [niche keywords]. Save findings to ~/research/inbox/.

2. Every morning at 7am: check arXiv for new papers in [topics]. Save to ~/research/inbox/.

3. Every day at 9am: monitor these competitor URLs for changes: [url1, url2, url3]. Use a wakeAgent script that hashes each page and only wakes you if content changed. Save diff summary to ~/research/inbox/.

4. Every monday at 8am: scan Product Hunt for AI launches from the past 7 days. Save top 10 to ~/research/inbox/.

Deliver confirmation to Telegram for all.“

Open Analyst and say:

“Set up two cron jobs:

1. Every day at 10am: check ~/research/inbox/ for new files. Use a wakeAgent script. If files found: feed them to NotebookLM for synthesis. Extract key insights. Write structured notes to Obsidian wiki. Tag confidence levels. Flag contradictions. Move processed files to ~/research/processed/.

2. Every friday at 3pm: run weekly deep synthesis. Review all wiki entries from this week. Identify patterns, trends, and gaps. Write a weekly synthesis note.

Deliver summary to Telegram for both.“

Open Briefer: same prompt as Standard.

Step 5 — Connect Telegram:

hermes dashboard → Channels → Telegram

@BotFather → /newbot → copy token → @userinfobot → copy user ID → paste both in Dashboard → restart gateway.

One bot, all three profiles deliver to the same chat.

Step 6 — Test:

Tell Scout: “run your web search cron now.” Wait for inbox file. Tell Analyst: “process the inbox.” Verify NotebookLM received the source. Check wiki entry. Next morning: check Telegram for brief.

WHAT THE MORNING LOOKS LIKE

8:00 AM. Telegram pings. Briefer delivers:

MORNING BRIEF — June 17, 2026

1. [verified] Competitor X updated pricing page. Removed free tier. Added enterprise plan at $299/mo. → review positioning against our offer today.

2. [likely] arXiv paper on agent memory consolidation aligns with our LLM Wiki approach. → read paper, consider wiki post about it.

3. [verified] Hermes v0.16.1 hotfix released. Dashboard reload fix + 3 security patches. → run hermes update on VPS.

4. [unverified] X thread claims 40% cost reduction with new DeepSeek model on agent workloads. → needs verification before posting about it.

5. [conflicting] two sources disagree on NotebookLM enterprise API pricing. → flagged in wiki contradictions folder.

Token spend this week: $4.20

You read 5 bullets. Decide what matters. Reply to Telegram if you want the agent to act on something. The research happened while you were asleep.

YanXbt@IBuzovskyi·Jun 14 ArticleHERMES AGENT BUILDS ITSELF WHILE YOU SLEEP. THE COMPLETE GUIDE TO THE 9-HOUR OVERNIGHT WORKFLOW.Most AI agents wait for you to type. Hermes Agent does not. And it does something none of the others do: it gets smarter overnight. Between 11PM and 8AM, a properly configured Hermes setup monitors…661519158K

The full 9-hour overnight workflow that powers this morning brief.

COST BREAKDOWN

Three pricing paths. Pick what fits your setup.

PATH 1 — NOUS PORTAL (simplest, one subscription)

One subscription covers all three profiles. 300+ models. Tool Gateway included (web search, image gen, TTS, browser automation). 10% off token-billed providers. Routes through OpenRouter under the hood. Cron jobs bill against your subscription automatically.

Setup:

hermes setup –portal

One OAuth. One login. Everything works.

Check portal.nousresearch.com for current tier pricing and credit amounts. Monitor with /usage and hermes portal info to track how fast credits deplete.

PATH 2 — OPENROUTER API (pay per token)

Pay per token. No subscription. Lowest cost. Requires managing OpenRouter API key but one key covers all models.

Best for: people who want minimum spend and don’t mind monitoring token usage closely.

PATH 3 — CHATGPT SUB + SONNET API (most generous tokens)

Scout and Briefer run on GPT-5.5 with generous included tokens from the $20 ChatGPT subscription. Analyst runs on Sonnet through a separate Anthropic API key.

Higher total cost but simplest token management. Scout and Briefer rarely hit limits.

COMPARISONS

For reference: a part-time research assistant costs $1,500-3,000/month.

All cost estimates based on typical usage patterns for this setup (~1.3M tokens/month across three profiles). Actual costs depend on cron frequency, synthesis depth, and model choice. Monitor with /usage.

DAY 1 → WEEK 2 → MONTH 1

DAY 1

→ Three profiles created, SOUL.md written → 4-5 cron jobs running (Scout scans, Briefer reports) → Empty Obsidian vault → Empty memory across all profiles → First morning brief: generic, broad, not personalized

Useful but not impressive. The system is cold.

WEEK 2

→ Scout has found 50-100 sources → Analyst has written 30-40 wiki entries → Obsidian graph shows first cross-references → Memory has accumulated niche context → Morning briefs reference YOUR specific projects → Analyst starts tagging patterns you missed

First moment: “this found something I wouldn’t have searched for.”

MONTH 1

→ 200+ wiki entries with cross-references → Contradictions flagged and tracked → Scout crons refined (you told it to stop finding certain things) → Analyst created 5-10 skills for recurring synthesis patterns → Briefer knows your priorities and goal cadence → Curator has pruned stale skills across all profiles → Morning brief feels like it was written by someone who knows your work

The system produces insights you did not ask for. That is the compounding.

LIMITATIONS

NotebookLM wrapper can break. notebooklm-mcp-cli uses browser automation. No official consumer API exists. If Google changes endpoints, the wrapper needs updating. Always have the fallback path in Analyst SOUL.md.

**Scout misses paywalled content.**Web search and X search hit public content only. Paywalled articles, private repos, gated communities are invisible to Scout. Add those sources manually to the inbox.

**Analyst can misclassify confidence levels.**The [verified]/[unverified] tagging depends on the model’s judgment. Hermes does not independently fact-check claims. Cross-reference important findings yourself before acting on them.

**Token costs scale with volume.**More Scout crons = more inbox files = more Analyst processing = more tokens. Start with 3-4 Scout crons. Add more only when you know the cost per cron run. Monitor with /usage.

**Human review still matters.**This is a research department, not an autopilot. The morning brief is a starting point. Decisions are yours. The system finds and organizes. You decide and act.

RELATED ARTICLES

• Hermes Agent as a Personal AI Operating System — full 17-layer architecture

• 8 Loops Inside Hermes Agent (And Why They Compound) — the loop architecture

• HERMES AGENT SOUL.MD: WHY 50 LINES MATTER MORE THAN YOUR MODEL — identity guide

• HERMES AGENT BUILDS ITSELF WHILE YOU SLEEP — the overnight workflow

• wakeAgent and no_agent cron flags — the $0 gate for monitoring

Subcribe to my SubStack: https://substack.com/@yanxbt for more articles

OFFICIAL SOURCES

• Profiles documentation

• LLM Wiki skill

• Cron jobs

• MCP servers

• Memory system

• notebooklm-mcp-cli

All technical details verified against Hermes Agent v0.16.0 documentation and referenced build guides.