OmniRoute is an AI gateway that gives coding agents one endpoint for accessing models from many providers, with quota-aware fallback between them. It is for people using Claude Code, Codex, Cursor, OpenCode, Cline, Copilot, and similar tools who want a single route to their available AI models.
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 diegosouzapw/OmniRoute --skill merge-batchgit clone --depth 1 https://github.com/diegosouzapw/OmniRouteWrote 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/diegosouzapw/omniroute/merge-batch)<a href="https://agentmods.dev/skills/diegosouzapw/omniroute/merge-batch"><img src="https://agentmods.dev/badge/skills/diegosouzapw/omniroute/merge-batch/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/diegosouzapw/omniroute/merge-batch"><img src="https://agentmods.dev/badge/skills/diegosouzapw/omniroute/merge-batch.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 | $0.00046 | $0.01113 |
| Opus 5.5 | $0.00018 | $0.00445 |
| Sonnet 5.5 | $0.00009 | $0.00223 |
| Haiku 4.5 | $0.00005 | $0.00111 |
Grade A, and why
merge-batch 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 today.
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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Merge batch: shared preparation, exact admission
Read ../_shared/base-green.md, ../_shared/validation-gate.md and
../_shared/merge-gates.md. Their current-candidate, authority, protected-surface
and credit rules govern this workflow. Historical throughput, administrative
merge and blanket rebaseline exceptions are superseded by that protocol.
Batching shares discovery and conflict diagnosis. It does not make a combined board's PASS evidence valid for different intermediate squash candidates.
1. Intake and authority
Resolve each selected PR's open/draft state, current head/base, target, contributor and write authority. Read the diff and current freeze state, and check active worktree ownership. Hold any item another session is working. A request to analyze or prepare a list is not authority to merge or close it.
Run scripts/quality/check-skill-contract.mjs against the actual installed skills
root before admission. Missing or drifted modeled instructions mean HOLD; a
template-only check does not verify the installation being used.
Show a 1–5 star table with rationale, overlaps and holds. Stars prioritize work; they do not waive required checks or protected-surface approval. Ask the owner before acting on low-star items or dispatching rework not already authorized. For a subsumed item, prove the current behavior and obtain the separate per-item closure decision; preserve the original PR when its work is being used.
Completion: selected scope, target, ownership and permitted actions are explicit.
2. Build the diagnostic board
Pin the approved live base and record the ordered PR heads. Use one owned branch
and worktree under .claude/worktrees/, with active hooks and dependencies
matching the candidate lockfile. Verify dependency identity instead of assuming a
copy from the shared checkout is current. Do not install over shared hard links.
Board the selected refs and verify each is an ancestor of the combined head. Resolve overlaps without dropping sibling behavior or contributor attribution. Keep unrelated worktrees and the shared checkout untouched; never use git stash. Parallel preparation needs authorized delegation and disjoint write ownership.
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.
- today First seen · 103 lines · 46 tokens per session scan A ceb23e339531
merge-batch is a skill published in the GitHub repository diegosouzapw/OmniRoute (77,818 stars, last pushed today), licensed MIT. It adds 46 tokens to every session and 1,113 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-10-18.
Other skills, from other repositories
babysit
Watch a pull request or review cycle until it is ready to merge. Use when asked to babysit, monitor, or keep checking PR comments, reviews, and CI until all actionable issues are resolved.
code-review
Review the current diff, or a PR number/branch/path target, for correctness bugs (plus reuse/simplification/efficiency cleanups where the model's review recipe covers them) at the given effort level (low|medium|high|xhigh|max: from few high-confidence findings up to many, some of them uncertain; ultra: deep…
github
GitHub via gh CLI: PRs, issues, reviews, repos, auth.
babysit
Watch an open PR — fix failing CI, handle the straightforward review comments, and drive it to a mergeable state. Claude Code analog of Cursor's built-in /babysit. Use after opening a PR when the user wants the agent to shepherd it without re-prompting.
make-pr-easy-to-review
Prepare PRs for review by cleaning noisy history, improving PR descriptions, and adding reviewer guidance without changing code behavior. Use for "make this easy to review", "tidy this PR", "clean up commits", or "annotate the diff".
pr-review
Review open PRs two ways - default is a per-PR deep review with severity-tagged findings, inline comments, and a verdict; --survey runs a risk-tiered triage digest of what's safe to merge first.