Hermes Agent is an AI assistant that learns from its use by creating and improving skills, retaining knowledge, searching past conversations, and adapting to its users. It is for people who want to run an agent through a terminal or messaging platforms while connecting it to different AI models and scheduled tasks.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add NousResearch/hermes-agent --skill unbrokergit clone --depth 1 https://github.com/NousResearch/hermes-agentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nousresearch/hermes-agent/unbroker)<a href="https://agentmods.dev/skills/nousresearch/hermes-agent/unbroker"><img src="https://agentmods.dev/badge/skills/nousresearch/hermes-agent/unbroker/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nousresearch/hermes-agent/unbroker"><img src="https://agentmods.dev/badge/skills/nousresearch/hermes-agent/unbroker.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Snyk warn
- NVIDIA SkillSpector warn
SkillSpector: 4 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 3 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Memory Poisoning · line 183 Skill injects content designed to persist in agent memory or context across interactions. Persistent injection can alter agent behavior long after the initial interaction.Fix: Do not allow untrusted input to persist in agent memory or context. Validate all content before storing and implement memory isolation between sessions.
- medium Excessive Agency · line 200 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 277 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00015 | $0.06299 |
| Opus 5 | $0.00008 | $0.03150 |
| Sonnet 5 | $0.00003 | $0.01260 |
| Haiku 4.5 | $0.00002 | $0.00630 |
Grade A, and why
unbroker scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 318 lines — stays where its author put it; the contents beside it link to each section on GitHub.
unbroker
Find where a person's personal information (name, addresses, phone, email, relatives) is exposed on data brokers and people-search sites, then remove it - automatically where possible, with guided human steps only where a site demands a CAPTCHA, government ID, phone call, or fax. Manages multiple people independently. It does not defeat anti-bot systems, does not act on anyone without recorded consent, and does not remove public records (voter/property/court) or accounts the person controls.
The Python CLI (scripts/pdd.py) owns the deterministic state - config, dossiers + consent, the
broker database, tier planning, the ledger, drafts, reports, email sending (SMTP), verification-link
polling (IMAP), and the autonomous action queue (next). You (the agent) do the scanning and
form-driving with native tools: web_extract and browser_navigate for searching and web forms, and
cronjob for recurring re-scans.
Autonomy contract
This skill is designed to run hands-off. After intake (+ recorded consent) there are exactly TWO
legitimate human touchpoints: (1) the intake conversation itself, and (2) ONE consolidated human-task
digest at the end of the run ($PDD tasks). Between those:
- Never ask the operator to choose configuration.
$PDD setup --autodetects capabilities and picks the most autonomous valid config itself. - Never pause before individual submissions when
autonomy=full(the default): the consent recorded at intake is standing authorization for T0-T2 opt-outs. (autonomy=assistedrestores per-submission confirmation for cautious operators - honorconfirm_firstflags innextoutput.) - Never interrupt the run for human-only work. Record it (
record ... human_task_queued --reason "...") and keep going; it all surfaces once in the final digest. - Drive the whole run as a loop over
$PDD next <subject>- it returns the exact ordered actions to take right now (scan, poll verification, re-check, opt out parents-first, requeue blocked), plus the human digest. Execute every action, record outcomes, re-runnext, repeat untildone_for_now. Then present the digest, report, and schedule the cron.
What ships with it
55 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- assets/unbroker.png 350 KB
- README.md 10 KB
- references/brokers/addresses.json 1.8 KB
- references/brokers/advancedbackgroundchecks.json 2.9 KB
- references/brokers/beenverified.json 3.5 KB
- references/brokers/clustal.json 2.3 KB
- references/brokers/clustrmaps.json 1.4 KB
- references/brokers/cyberbackgroundchecks.json 1.4 KB
- references/brokers/familytreenow.json 1.4 KB
- references/brokers/fastpeoplesearch.json 2.2 KB
- references/brokers/intelius.json 11 KB
- references/brokers/mylife.json 974 B
- references/brokers/nuwber.json 1.4 KB
- references/brokers/peekyou.json 1.4 KB
- references/brokers/peoplefinders.json 1.5 KB
- references/brokers/radaris.json 3.4 KB
- references/brokers/rehold.json 2.1 KB
- references/brokers/searchpeoplefree.json 1.4 KB
- references/brokers/socialcatfish.json 2.1 KB
- references/brokers/spokeo.json 3.1 KB
- references/brokers/thatsthem.json 2.5 KB
- references/brokers/truepeoplesearch.json 2.7 KB
- references/brokers/usphonebook.json 1.4 KB
- references/brokers/whitepages.json 4.0 KB
- references/legal/ccpa.md 1.1 KB
- references/legal/drop.md 2.1 KB
- references/legal/gdpr.md 737 B
- references/methods.md 27 KB
- references/site-playbooks.md 5.1 KB
- references/state-machine.md 3.2 KB
- scripts/autopilot.py 21 KB runs code
- scripts/badbool.py 6.0 KB runs code
- scripts/brokers.py 2.6 KB runs code
- scripts/cdp.py 5.8 KB runs code
- scripts/config.py 6.0 KB runs code
- scripts/crypto.py 3.0 KB runs code
- scripts/dossier.py 4.8 KB runs code
- scripts/email_modes.py 3.0 KB runs code
- scripts/emailer.py 14 KB runs code
- scripts/ledger.py 7.8 KB runs code
- scripts/legal.py 2.3 KB runs code
- scripts/paths.py 2.1 KB runs code
- scripts/pdd.py 46 KB runs code
- scripts/registry.py 14 KB runs code
- scripts/report.py 7.0 KB runs code
- scripts/scan.py 1.3 KB runs code
- scripts/storage.py 4.3 KB runs code
- scripts/tiers.py 14 KB runs code
- scripts/vectors.py 2.1 KB runs code
- templates/consent/authorization.md 742 B
- templates/emails/ccpa-authorized-agent.txt 792 B
- templates/emails/ccpa-deletion.txt 686 B
- templates/emails/ccpa-indirect-deletion.txt 1.4 KB
- templates/emails/gdpr-erasure.txt 606 B
- templates/emails/generic-optout.txt 470 B
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 318 lines · 15 tokens per session scan A 781ddef02363
unbroker is a skill published in the GitHub repository NousResearch/hermes-agent (243,598 stars, last pushed today), licensed MIT. It adds 15 tokens to every session and 6,299 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
rubyllm
Build and maintain Ruby or Rails applications with the RubyLLM AI framework. Use for chats, agents, tools, structured output, media generation, transcription, OCR, moderation, embeddings, reranking, Rails integration, and RubyLLM upgrades; not for contributing to the framework itself.
67-agency-vendor-brief
Dung khi thue NGUOI NGOAI lam — agency, freelancer, production house, vendor: scope of work, spec deliverable, so vong sua, quy trinh duyet, moc thanh toan, dieu khoan IP va bao mat, bang danh gia vendor sau du an. Kich hoat khi user nhac 'brief agency', 'thue freelancer', 'quan ly vendor', 'scope of work', 'agency…
standards-expert
Expert-level ISO standards, quality management, compliance, and certification. Use when the user mentions ISO standards, quality management, compliance, or certification, or when the task involves Compliance & Certification, Auditing, or Continuous Improvement.
lawyer-expert
Expert-level legal systems, contracts, compliance, and legal technology. Use when the user mentions legal, contracts, compliance, law, or legal tech, or when the task involves Legal Systems, Legal Technology, Compliance Frameworks, or Contract Management.
banking-expert
Expert-level banking systems, core banking, regulations, and banking technology. Use when the user mentions finance, core banking, regulations, or payments, or when the task involves Banking Systems or Key Technologies.
legal-research
Use when the user asks anything about German or European law — statutes (BGB, StGB, AO, EStG, KSchG, AGG, …), EU regulations/directives (GDPR, AI Act, MiCA, NIS2, DSA, DMA, …), federal court decisions (BGH, BVerfG, BAG, BSG, BFH, BPatG), CJEU judgments, or legal-compliance questions in code or documents. Triggers on…