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 agentmods add skills/tweakoz/orkid/obtnetnpx skills add tweakoz/orkid --skill obtnetgit clone --depth 1 https://github.com/tweakoz/orkidWrote 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/tweakoz/orkid/obtnet)<a href="https://agentmods.dev/skills/tweakoz/orkid/obtnet"><img src="https://agentmods.dev/badge/skills/tweakoz/orkid/obtnet.svg" alt="Measured on agentmods" 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 | $0.00055 | $0.01742 |
| Opus 5 | $0.00028 | $0.00871 |
| Sonnet 5 | $0.00011 | $0.00348 |
| Haiku 4.5 | $0.00006 | $0.00174 |
Grade A, and why
obtnet 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 2d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
obtnet — one controller, many nodes
Every command prints ONE greppable verdict line last: [obtnet] ok|FAIL <verb> <node> ....
Trust it. Do not fetch or dump logs on success. --json on any verb for tool parsing.
Controller resolution: --controller > $OBTNET_CONTROLLER > ~/.obt-global/obtnet.json
(set once with obt.net.py config tcp://HOST) > localhost. Bare HOST or HOST:PORT are
accepted anywhere an address is (port defaults to 7461).
Node names and selectors
obt.net.py list shows the fleet. Any <node> argument accepts a literal name or a
@ selector, routed to the least-busy live match:
@any · @linux / @mac · @gpu · @gpu=5090 (capability substring) · @5090
(substring across name/os/arch/gpu) · comma-composed: @linux,gpu.
FAN-OUT on run/build/test/scene/sync/gitsync: @each (every live node), @each:linux
(every match), or a,b lists — execution verbs submit to ALL then wait (simultaneous
cross-machine builds/tests), working-copy verbs loop; per-node verdicts + a fleet summary.
One happy path per task
obt.net.py list # fleet + load + clock offsets
obt.net.py run <node> [--timeout S] [--cwd D] -- CMD... # short sync command (<~60s)
obt.net.py submit <node> [--out GLOB] -- CMD... # long work -> job id
obt.net.py build <node> [--cwd D] # ork.build.py (or -- CMD);
# verdict: errors=N warnings=N
# (error-grep beats lying rc0)
obt.net.py test <node> -- CMD... # verdict: passed=N failed=N
obt.net.py scene <node> [--out G] [--env K=V] -- CMD... # OFFSCREEN forced unless
# --windowed (owner consent);
# cook/settle/fps lines surfaced
obt.net.py wait <node> j-XXXX # verdict; auto stderr-tail on fail
obt.net.py log <node> j-XXXX [--stderr --tail N --grep P] # bounded peek at a LIVE job
obt.net.py jobs # fleet-wide job table
obt.net.py cancel <node> j-XXXX # SIGTERM the job's pgid
obt.net.py fetch <node> <sha256> <dest> # artifact by sha (from verdicts)
obt.net.py sync <node> <local_dir> <remote_dir> # make remote == local (git-free)
obt.net.py sync <node> L R --pull # make local == remote
obt.net.py gitsync <node> <local_repo> <remote_repo> # align git base (branch+HEAD)
# via bundle, NO push; --tree
# chains the uncommitted delta
# LFS objects the node lacks
# ride along (--no-lfs off,
# --lfs-max-mb caps a seed)
obt.net.py diff <node> <local_dir> <remote_dir> # +/-/M/L classified; rc0=equal
obt.net.py watch [--node N] [--grep P] # live fleet event tail (ctrl-c)
Nodes run inside OBT shells, so staged tools (ork.build.py, obt.dep.*, ork.cpp.db.*)
work directly: obt.net.py run @linux,gpu -- ork.cpp.db.search.py MySymbol --porcelain.
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.
- 2d ago First seen · 106 lines · 55 tokens per session scan A 323faa13df23
obtnet is a skill published in the GitHub repository tweakoz/orkid (35 stars, last pushed 22d ago), licensed MIT. It adds 55 tokens to every session and 1,742 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 2026-09-01.
Other skills, from other repositories
host-pattern
Use when adding a new domain to Nuclear's plugin system, or implementing a host. Covers the host pattern (how player functionality is exposed to plugins), the host interface and API class structure, how hosts are implemented in the player, error handling conventions, and what files to create and modify. Trigger…
creating-components
Use when creating new UI components in packages/ui. Covers component structure, tests, stories, and what to avoid.
writing-docs
Use when writing or editing documentation in packages/docs. Covers Gitbook markdown syntax, special blocks, page structure, and the SUMMARY.md table of contents. Trigger phrases include "write docs", "add documentation", "docs page", "gitbook", "user manual".
writing-plugins
Use when writing, scaffolding, or modifying Nuclear plugins. Covers plugin structure, manifest, entry point, provider types, available APIs, and publishing. Trigger phrases include "create a plugin", "write a plugin", "plugin scaffold", "streaming provider", "metadata provider".
analyzing-linux-elf-malware
Analyze malicious Linux ELF binaries — botnets, cryptominers, ransomware, and rootkits targeting Linux servers, containers, and cloud infrastructure — through static analysis, dynamic tracing, and reverse engineering of x8664 and ARM samples. Use when investigating Linux malware, triaging a suspicious ELF binary…
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.