Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/skydoves/compose-performance-skillsnpx agentmods add skills/skydoves/compose-performance-skills/tracing-recompositions-at-runtimeWrote 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/skydoves/compose-performance-skills/tracing-recompositions-at-runtime)<a href="https://agentmods.dev/skills/skydoves/compose-performance-skills/tracing-recompositions-at-runtime"><img src="https://agentmods.dev/badge/skills/skydoves/compose-performance-skills/tracing-recompositions-at-runtime/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/skydoves/compose-performance-skills/tracing-recompositions-at-runtime"><img src="https://agentmods.dev/badge/skills/skydoves/compose-performance-skills/tracing-recompositions-at-runtime.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00215 | $0.04241 |
| Opus 5.5 | $0.00086 | $0.01696 |
| Sonnet 5.5 | $0.00043 | $0.00848 |
| Haiku 4.5 | $0.00021 | $0.00424 |
Grade A, and why
tracing-recompositions-at-runtime 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 — 304 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tracing Recompositions at Runtime — @TraceRecomposition, logcat, and the live heatmap
Layout Inspector counts recompositions and surfaces Argument Change Reasons, but it works only in debug, where Live Literals and the interpreted Compose runtime inflate counts. @TraceRecomposition from skydoves/compose-stability-analyzer instruments a composable at compile time and emits per-recomposition diffs (which state changed, what value transition) to logcat under the Recomposition tag. The instrumentation works in any build the developer enables it for — including release-with-debug-symbols — and feeds the IntelliJ / Android Studio plugin's live recomposition heatmap.
This skill is the release-mode complement to ../../recomposition/debugging-recompositions/SKILL.md. Layout Inspector is debug-only, fast to set up, and good for the first triage. @TraceRecomposition is the ground-truth confirmation: instrument the suspect composable, ship a release+R8 build of the dev APK, run the user journey, and read the per-recomposition log lines.
When to use this skill
- A composable recomposes more than expected and Layout Inspector counts are inconclusive (the count differs between debug and release, or the suspect is an inline composable not covered by Layout Inspector).
- A stability or strong-skipping fix needs to be confirmed against a release-equivalent build before merging.
- A developer wants per-state-and-per-parameter change diffs printed inline rather than clicking through the Layout Inspector tree.
- A team wants to baseline a composable's recomposition count for an SLO ("PriceTicker recomposes ≤ once per price update; never per parent tick").
- The user mentions
@TraceRecomposition, "trace recomposition", "compose-stability-analyzer", "recomposition logcat", "recomposition heatmap", or "release-mode recomposition".
When NOT to use this skill
- The developer just wants recomposition counts in debug — Layout Inspector is faster to set up. See
../../recomposition/debugging-recompositions/SKILL.md. - The build is a production release with no diagnosis intent — the instrumentation must be gated off. See the runtime-toggle pattern below.
- The team needs a CI gate that fails on future stability regressions — runtime tracing is for diagnosis; gating is
../../stability/enforcing-stability-in-ci/SKILL.md(stabilityCheck). - The need is per-frame timing or end-to-end user-perceived perf, not recomposition counts — that is
../generating-baseline-profiles/SKILL.mdwithMacrobenchmarkRule+FrameTimingMetric.
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 · 304 lines · 215 tokens per session scan A 1fe2d29e9a99
tracing-recompositions-at-runtime is a skill published in the GitHub repository skydoves/compose-performance-skills (520 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 215 tokens to every session and 4,241 once invoked, about $0.0009 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 2026-08-30.
Other skills, from other repositories
compose-performance
Use when investigating Jetpack Compose recomposition cost, compiler stability reports, skippability, unstable parameters, frame-rate State reads, cross-phase snapshot back-writing, or @ReadOnlyComposable contracts.
media-playback-footguns
Media player traps on Media3/ExoPlayer and desktop engines: crossfade, audio focus, fades, loudness, DSP, queues and shuffle, position restore, service lifecycle, group listening. Use when playback state, queue or audio levels go wrong.
webview-footguns
WebView traps across device classes: startup that freezes low-RAM phones, a killed renderer that takes the app with it, devices with no usable provider, AndroidView lifecycle and cookies. Use when a web screen freezes, vanishes or crashes only on some devices.
claude-android-ninja
Build and migrate Android apps with Kotlin, Jetpack Compose, MVVM, Hilt, Room 3 (KSP, SQLiteDriver, Flow/suspend DAOs), Navigation3, and multi-module Gradle. Use for new projects or modules, Compose screens and ViewModels, Room 3 and RemoteMediator, API 37 / targetSdk migration, Play Integrity client wiring…
compose-stability-diagnostics
Use when writing or reviewing Jetpack Compose parameter stability, compiler reports, skippability, unstable UI state classes, collection parameters, or Kotlin 2.0+ strong skipping behavior.
compose-recomposition-performance
Use when investigating Jetpack Compose recomposition performance, skippable/restartable composables, composables.txt or compiler reports, Layout Inspector recomposition counts, back-writing snapshot state across phases, or frame-rate State reads in composition vs layout/draw, and it is not yet clear whether the cause…