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 existential-birds/beagle --skill combine-code-reviewgit clone --depth 1 https://github.com/existential-birds/beagleWrote 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/existential-birds/beagle/combine-code-review)<a href="https://agentmods.dev/skills/existential-birds/beagle/combine-code-review"><img src="https://agentmods.dev/badge/skills/existential-birds/beagle/combine-code-review/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/existential-birds/beagle/combine-code-review"><img src="https://agentmods.dev/badge/skills/existential-birds/beagle/combine-code-review.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 | $0.00043 | $0.00850 |
| Opus 5.5 | $0.00017 | $0.00340 |
| Sonnet 5.5 | $0.00009 | $0.00170 |
| Haiku 4.5 | $0.00004 | $0.00085 |
Grade A, and why
combine-code-review scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Combine Code Review
Quick Reference
| Issue Type | Reference |
|---|---|
| Publishers, Subjects, AnyPublisher | references/publishers.md |
| map, flatMap, combineLatest, switchToLatest | references/operators.md |
| AnyCancellable, retain cycles, [weak self] | references/memory.md |
| tryMap, catch, replaceError, Never | references/error-handling.md |
Review Checklist
- All
sinkclosures use[weak self]when self owns cancellable - No
assign(to:on:self)usage (useassign(to: &$property)or sink) - All AnyCancellables stored in Set or property (not discarded)
- Subjects exposed as
AnyPublisherviaeraseToAnyPublisher() -
flatMapused correctly (not whenmap + switchToLatestneeded) - Error handling inside
flatMapto keep main chain alive -
tryMapfollowed bymapErrorto restore error types -
receive(on: DispatchQueue.main)before UI updates - PassthroughSubject for events, CurrentValueSubject for state
- Future wrapped in Deferred when used with retry
When to Load References
- Reviewing Subjects or publisher selection → publishers.md
- Reviewing operator chains or combining publishers → operators.md
- Reviewing subscriptions or memory issues → memory.md
- Reviewing error handling or try* operators → error-handling.md
Hard gates (before you report findings)
Complete in order. Do not skip ahead while a prior gate is open.
- Scope — Pass: You name at least one file or type under review that imports Combine or uses APIs from the Quick Reference (e.g.
AnyPublisher,@Published,PassthroughSubject). If none apply, stop with “out of scope.” - Subscription retention — Pass: For each
sink,assign, andstore(in:)in scope, you state where theAnyCancellableis retained (property,Set, task lifetime) or mark ephemeral with a one-line reason (e.g. synchronous one-shot that cannot outlive caller). If you cannot tell from the snippet, say unknown and ask for surrounding storage, do not assume safe. - Retain-cycle claim — Pass: Confirmed leak findings state the capture chain (e.g. self → stored cancellable → closure strongly capturing self). Label suspected cases risk / verify, not confirmed leaks. When arguing safety, cite
[weak self],[unowned self], or non-capturing patterns you relied on. - UI / main thread — Pass: For updates to UIKit/SwiftUI from a chain, you either point to
receive(on: DispatchQueue.main),@MainActor, or equivalent before the UI work, or flag missing scheduling withfile:line. - Severity and checklist — Pass: Every high or critical item includes
file:line(or exact pasted lines) and names which Review Checklist row it breaks. Lower-severity notes may omit line numbers but must still be reproducible from named files.
What ships with it
4 files 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.
- 1mo ago First seen · 54 lines · 43 tokens per session scan A 82d3d68d3865
combine-code-review is a skill published in the GitHub repository existential-birds/beagle (82 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 43 tokens to every session and 850 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 2026-09-16.
Other skills, from other repositories
axiom-audit-codable
Use when the user mentions Codable review, JSON encoding/decoding issues, data serialization audit, or modernizing legacy code.
axiom-swift-simplifier
Use when the user wants to simplify Swift code, reduce boilerplate, or make Swift more readable and idiomatic without changing behavior.
kotlin-concurrency-and-flow
Use when writing or reviewing Kotlin coroutine scope ownership, raw Thread or Executor work, init launches, non-suspending launch APIs, runBlocking, cancellation, StateFlow, SharedFlow, Channel, stateIn, SharingStarted, state updates, or one-shot events.
kotlin-api-design
Use when designing or reviewing Kotlin function ownership, member or extension functions, factories, single-field domain types, value classes, data classes, Kotlin Multiplatform expect/actual declarations, or platform service boundaries.
coding-best-practices
Reviews Swift/iOS code for adherence to modern Swift idioms, Apple platform best practices, architecture patterns, and code quality standards. Use when user mentions best practices, code review, clean code, refactoring, or wants to improve code quality.
swift-concurrency-review
Review Swift 6 strict-concurrency and SwiftUI code for idiom and build-breaking issues - non-Sendable across actor boundaries, @MainActor witness vs nonisolated protocol requirements, Combine/ObservableObject usage, force-unwraps, #Predicate macro limits, the 6.3.x Binding IRGen crash, missing #if os() guards, and…