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 ShaftHQ/SHAFT_ENGINE --skill omniroutegit clone --depth 1 https://github.com/ShaftHQ/SHAFT_ENGINEWrote 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/shafthq/shaft_engine/omniroute)<a href="https://agentmods.dev/skills/shafthq/shaft_engine/omniroute"><img src="https://agentmods.dev/badge/skills/shafthq/shaft_engine/omniroute.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 5.1 | $0.00022 | $0.03995 |
| Opus 5 | $0.00011 | $0.01997 |
| Sonnet 5 | $0.00004 | $0.00799 |
| Haiku 4.5 | $0.00002 | $0.00400 |
Grade A, and why
omniroute scanned grade A with 1 finding 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 3d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
curl -sf --max-time 2 http://127.0.0.1:20128/api/health How it starts
The opening of the file, as written. The whole thing — 289 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OmniRoute
Optional provider-neutral transport. It is not a workflow owner; select the
canonical workflow in execution workflows
first. Missing, stopped, unauthenticated, exhausted, or unqualified OmniRoute
is normal: use a qualified native implementer or SOLO.
Do not install OmniRoute, create provider accounts, or write operator
credentials. Do not expose credentials, account data, prompts, or consumer
code outside the approved bounded task. Receipts and repository files never
persist route, model, or provider IDs; live stdout of candidates may name
them for the current dispatch only.
Ensure the local gateway
Dashboard: http://127.0.0.1:20128/home. Health (anonymous JSON
status/timestamp is enough here):
command -v omniroute
curl -sf --max-time 2 http://127.0.0.1:20128/api/health
If omniroute is missing, do not install it; use native host-session models.
If the binary exists and health fails, start loopback only:
OMNIROUTE_SERVER_HOST=127.0.0.1 omniroute serve --port 20128 --no-open
Never bind a non-loopback address. Never use a remote --base-url.
Live catalog on every dispatch
Do not keep a session catalog file and do not cache positive catalog rows.
Remaining tokens change after each delegate, so query live models and
usage quota immediately before every dispatch. A user-local
provider-exhaustion backoff cache (XDG state, never committed) may store only
negative exhausted_until / retry-after entries; it is not a positive catalog cache.
Update it on exhaustion signals (state==exhausted, remaining<=0, HTTP 429,
quota-reset / insufficient-balance / stream-disconnect text). Pass the failed
identity through candidates(..., diagnostic=..., failed_identity_sha256=..., failed_provider=...) so the next ranking skips that entry until expiry.
omniroute --output json models
omniroute --output json usage quota
omniroute --output json models <provider>
python3 chaos-engine/skills/omniroute/scripts/runner.py candidates --capability mechanical|default|most-intelligent
What ships with it
1 file 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.
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.
- 3d ago First seen · 289 lines · 22 tokens per session scan A b165940e9622
omniroute is a skill published in the GitHub repository ShaftHQ/SHAFT_ENGINE (408 stars, last pushed today), licensed MIT. It adds 22 tokens to every session and 3,995 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
flowproof-config
Configure flowproof's SAP GUI, Fiori, and AI authoring credentials by walking the user through flowproof config sap / flowproof config fiori / flowproof config ai. Use when the user wants to set up, change, or check their SAP/Fiori login or model authoring key, or when a flow run fails because SAPUSER, FIORIPASSWORD…
java-coding-standards
A set of Java coding standards for Spring Boot services, covering naming, immutable data, optional values, streams, and exceptions.
RBT Manual Testing
Skill sinh manual test cases với 2 modes — QUICK (sinh nhanh từ requirements) và FULL RBT (quy trình AI-RBT 6 bước có đánh giá rủi ro). Master skill cho mọi tác vụ manual test case.
Framework Architect
Skill thiết kế và scaffold automation framework hoàn chỉnh cho Playwright, Selenium, và Appium — bao gồm project structure, base classes, config management, reporting, và CI/CD integration.
UI Debug Agent
Skill inspect web/mobile applications bằng browser tools, phân tích DOM elements, xác định locators ổn định, debug UI automation failures, và hỗ trợ sinh Page Object classes.
jira_integration
Skill tích hợp Jira/Xray — lấy requirements từ Jira, xác thực Xray, và đẩy kết quả test lên Xray Cloud/Server.