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 env-safe-executiongit 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/env-safe-execution)<a href="https://agentmods.dev/skills/milkyway008/hermes-otg/env-safe-execution"><img src="https://agentmods.dev/badge/skills/milkyway008/hermes-otg/env-safe-execution/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/env-safe-execution"><img src="https://agentmods.dev/badge/skills/milkyway008/hermes-otg/env-safe-execution.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 | $0.00055 | $0.01614 |
| Opus 5 | $0.00028 | $0.00807 |
| Sonnet 5 | $0.00011 | $0.00323 |
| Haiku 4.5 | $0.00006 | $0.00161 |
Grade A, and why
env-safe-execution 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 9d 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 — 127 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Env-Safe Execution — Check the Env Before You Execute
TRIGGER: about to run ANY executable that reads environment variables to resolve paths
or config — Hermes updaters/installers (hermes-setup.exe --update, install.ps1,
hermes update), package installers, SDK CLIs, clone/reclone scripts — ESPECIALLY when a
host install and a portable/OTG instance coexist on the same machine, or when working from
inside the OTG on host-machine tasks.
The Core Lesson
An agent running an executable from inside a portable/isolated environment is NOT the same
as the user running it. The user's shell is a clean host env; the agent's shell is a
contaminated portable env. Every env var the wrapper exported (HERMES_HOME, *_API_KEY,
install paths, ...) is inherited by EVERY subprocess the agent spawns — and any env-sensitive
app (updaters, installers, CLIs, SDKs) can be silently redirected to the wrong target.
| User's shell | Agent's shell (from inside OTG/portable) | |
|---|---|---|
| Env | clean host env | portable env: HERMES_HOME=E:\..., *_API_KEY, install paths |
hermes-setup.exe --update |
targets the host install | CAN target the OTG install |
The 5 Rules
Rule 1 — CHECK THE ENV BEFORE EXECUTING
Before running ANY command that touches another install/app, inspect what the subprocess would inherit:
cmd /c set HERMES_HOME
env | grep -i hermes
# or the target app's key vars: env | grep -iE 'API_KEY|INSTALL|HOME'
Know the leak before triggering it. If the command's target app reads env vars for paths/config, you MUST know what those vars are and what the subprocess will see.
Rule 2 — CLEAN-ENV OR EXPLICIT-TARGET
Scrub the portable env vars, or set them EXPLICITLY to the intended target, before running:
# scrub the portable vars (git-bash):
env -u HERMES_HOME -u HERMES_OTG -u TERMINAL_CWD <command>
# explicit target (cmd):
set "HERMES_HOME=%LOCALAPPDATA%\hermes" && <command>
NEVER rely on inherited values when the target is a different install.
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.
- 9d ago First seen · 127 lines · 55 tokens per session scan A 582a5ab40e3d
env-safe-execution is a skill published in the GitHub repository MilkyWay008/Hermes-OTG (15 stars, last pushed 28d ago), licensed MIT. It adds 55 tokens to every session and 1,614 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 2026-09-03.
Other skills, from other repositories
acquiring-disk-image-with-dd-and-dcfldd
Create forensically sound bit-for-bit disk images with dd or dcfldd on a Linux forensic workstation, preserving evidence integrity through hash verification (MD5/SHA) during acquisition. Use when imaging a suspect drive, USB device, or memory card for investigation, preserving volatile disk evidence during incident…
analyzing-linux-kernel-rootkits
Detect kernel-level rootkits in Linux memory dumps using Volatility3 linux plugins (checksyscall, lsmod, hiddenmodules), rkhunter system scanning, and /proc vs /sys discrepancy analysis to identify hooked syscalls, hidden kernel modules, and tampered system structures.
implementing-cloud-trail-log-analysis
Implementing AWS CloudTrail log analysis for security monitoring, threat detection, and forensic investigation using Athena, CloudWatch Logs Insights, and SIEM integration to identify unauthorized access, privilege escalation, and suspicious API activity.
performing-cloud-log-forensics-with-athena
Uses AWS Athena to query CloudTrail, VPC Flow Logs, S3 access logs, and ALB logs for forensic investigation. Covers CREATE TABLE DDL with partition projection, forensic SQL queries for detecting unauthorized access, data exfiltration, lateral movement, and privilege escalation. Use when investigating AWS security…
subdomain-enumeration
Map subdomains via crt.sh and subfinder at recon kickoff.
hunt-django
Hunt Django-specific vulnerabilities: DRF permission gaps, ORM injection, and admin exploitation.