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 MilkyWay008/Hermes-OTG --skill otg-pipgit clone --depth 1 https://github.com/MilkyWay008/Hermes-OTGWrote 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/milkyway008/hermes-otg/otg-pip)<a href="https://agentmods.dev/skills/milkyway008/hermes-otg/otg-pip"><img src="https://agentmods.dev/badge/skills/milkyway008/hermes-otg/otg-pip/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/milkyway008/hermes-otg/otg-pip"><img src="https://agentmods.dev/badge/skills/milkyway008/hermes-otg/otg-pip.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00059 | $0.01190 |
| Opus 5 | $0.00030 | $0.00595 |
| Sonnet 5 | $0.00012 | $0.00238 |
| Haiku 4.5 | $0.00006 | $0.00119 |
Grade A, and why
otg-pip 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 12d 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 — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Installing Python Packages for Hermes on OTG (otg-pip)
Trigger: Hermes on an OTG (portable/USB) package needs a Python package that
isn't available — module not found / ImportError, a feature that needs pip
(web search via ddgs, extraction via trafilatura, PDF/SSH/compression
libraries, ...), or a package that should have lazy-installed but didn't.
THE KEY FACT (read this first)
The frozen Hermes-OTG.exe cannot lazy-install packages. Hermes' native
allow_lazy_installs silently fails on OTG: sys.executable is the
PyInstaller bootloader (not a real python), and uv is not bundled. So any
optional module must be installed MANUALLY into the INTERNAL venv:
<OTG_ROOT>\data\hermes-agent\venv\
This venv is Python 3.11 — it matches the frozen exe's ABI. Its
site-packages is appended to the frozen agent's sys.path at startup
(Option B runtime hook in rthook_otg.py), so ANYTHING installed there is
directly importable by Hermes. fix-otg-paths keeps its pyvenv.cfg home
pointing at the bundled 3.11 base (dependencies\python-311) on every launch.
Resolve <OTG_ROOT>: dirname "$HERMES_HOME" — HERMES_HOME is
<OTG_ROOT>\data\ (note the trailing backslash).
HARD RULES
- ALWAYS install into the INTERNAL venv with its own tools:
<OTG_ROOT>\data\hermes-agent\venv\Scripts\pip.exe(or<OTG_ROOT>\data\hermes-agent\venv\Scripts\python.exe -m pip). - NEVER use the host machine's pip/python — it installs into the HOST environment, not the OTG package, and breaks portability.
- NEVER install into the EXTERNAL venv (
dependencies\python\.venv, Python 3.12) — that venv belongs to MCP servers and runs in SEPARATE processes. (MCP server deps go through theotg-mcpskill instead.) - Compiled packages (
.pyd) MUST be cp311-compatible — they import into the frozen 3.11 exe. Pure-Python wheels (py3-none-any) always work;cp312wheels will crash the agent with an ABI error. - Remember the append-only world: the frozen PYZ modules always win over
the internal venv. Don't try to "upgrade" a module that ships frozen
(e.g.
httpx,pydantic) — it won't take effect in the frozen process.
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.
- 12d ago First seen · 86 lines · 59 tokens per session scan A 97c80be38c00
otg-pip is a skill published in the GitHub repository MilkyWay008/Hermes-OTG (15 stars, last pushed 27d ago), licensed MIT. It adds 59 tokens to every session and 1,190 once invoked, about $0.0003 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 2026-08-31.
Other skills, from other repositories
flask-werkzeug-attack
Exploit Flask/Werkzeug debugger exposure for traceback and SECRET leaks.
python-debugpy
Debug Python with pdb, breakpoint(), post-mortem inspection, and debugpy remote or headless attach.
style-guide
Dedalus style guide quick reference. Key rules for writing code that lasts. Use when writing or reviewing Python code.
column-aligned-fields
Enforce column-aligned inline comments on dataclass and enum fields. Use when writing or reviewing Python dataclasses, enums, TypedDicts, or any structured type with inline field comments.
python-debugpy
Debug Python: pdb REPL + debugpy remote (DAP).
pytorch-lightning
High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.