![Abstract illustration of a modular Kotlin and Jetpack Compose toolkit](docs/assets/skills-header.webp)

# Skills

A set of skills for Kotlin, Jetpack Compose, Android development, and grounded
writing.

The repository is also a portable [Agent Plugins](https://agent-plugins.org/)
v1.0.0 package. Conforming clients discover the root [`plugin.json`](plugin.json)
and the immediate skill directories under [`skills/`](skills/).

## Install

With the [skills CLI](https://skills.sh):

```
npx skills add chrisbanes/skills
```

Or install as a Claude Code plugin:

```
/plugin marketplace add chrisbanes/skills
/plugin install chrisbanes-skills@chrisbanes-skills
```

Or install as a Codex plugin:

```
codex plugin marketplace add chrisbanes/skills --ref main
codex plugin add chrisbanes-skills@chrisbanes-skills
```

Or install as an OpenCode V2 plugin (2.0.22 or newer):

```json
{
  "plugins": ["chrisbanes-skills@git+https://github.com/chrisbanes/skills.git"]
}
```

OpenCode V1 1.18.29 or newer uses `plugin` instead of `plugins`.
See [`.opencode/INSTALL.md`](.opencode/INSTALL.md) for details.

### External skill providers

Most skills in this repository are self-contained. The workflows below compose
with skills from other repositories; installing `chrisbanes/skills` does not
install them, and the workflows never install them implicitly.

| Consumer | Requirement | Provider | Source and install |
|---|---|---|---|
| `deliver-spec` GitHub readiness | Conditional | `triage` | [Matt Pocock's skills](https://github.com/mattpocock/skills): `npx skills add mattpocock/skills --skill triage` |
| [`implement-with-subagents`](skills/implement-with-subagents/SKILL.md) behavior tasks | Conditional | `tdd` at approved test seams | [Matt Pocock's skills](https://github.com/mattpocock/skills): `npx skills add mattpocock/skills --skill tdd` |
| [`run-github-project`](skills/run-github-project/SKILL.md) explicit/repository-required test-first work | Conditional | `tdd` | [Matt Pocock's skills](https://github.com/mattpocock/skills): `npx skills add mattpocock/skills --skill tdd` |
| `run-github-project` Todo triage | Conditional | `triage` | [Matt Pocock's skills](https://github.com/mattpocock/skills): `npx skills add mattpocock/skills --skill triage` |
| `run-github-project` Wayfinder lane | Conditional | `wayfinder`, `research` | [Matt Pocock's skills](https://github.com/mattpocock/skills): use `--skill wayfinder` or `--skill research`; see the [workflow provider matrix](skills/run-github-project/references/workflow-providers.md) |

Delivery and parallel implementation use one fresh independent read-only reviewer
against the approved source and repository standards; no external review skill
or issue-tracker setup is required. Reviews include the production paths and
boundary cases examined and material uncertainty. The authoritative
[behavioral review and repair procedure](references/behavioral-review.md) covers
related boundary repairs, sensitive fixture calibration before expansion,
rendering-consumer tracing, repeated external findings and browser diagnostics.
`npm run references:sync` supplies identical local copies to its consumers;
edit the authoritative source, not the copies. The same mechanism supplies
[to-plan's boundary-validation rule](skills/to-plan/references/boundary-validation.md)
to implementation consumers.

Delivery and planning select affected existing tests through shared-structure
consumers and structural audits. Unsettled fixtures establish one sensitive case
before expanding the acceptance matrix; established fixtures reuse valid evidence.

Review mode in `implement-with-subagents`, and review or setup mode in
`run-github-project`, do not require these external skills. See the
[`run-github-project` provider matrix](skills/run-github-project/references/workflow-providers.md)
for its lane-specific fallback and blocking behavior.

## Skills

### Start here

- Working on Compose state or effects? Start with [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md).
- Investigating recomposition, stability, or jank? Start with [`compose-performance`](skills/compose-performance/SKILL.md).
- Comparing Android benchmark configurations or a measured Android default? Start with [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md).
- Reviewing Flow or coroutine architecture? Start with [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md).

### Routing

- [`using-chrisbanes-skills`](skills/using-chrisbanes-skills/SKILL.md) — route Kotlin and Jetpack Compose work to focused skills, adding a second only for an independent decision in the same change.

### Benchmarking

- [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md) — compare physical Android benchmark configurations with verified coverage, controlled conditions, trace-backed diagnosis of unstable rankings, and bounded conclusions.

### Jetpack Compose

#### State and side effects

- [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) — decide state ownership and effect lifecycle for local UI state, screen state holders, Flow collection, callbacks, cleanup, navigation, snackbar, analytics, and focus requests.

#### Performance

- [`compose-performance`](skills/compose-performance/SKILL.md) — diagnose stability, deferred reads, composition contracts, and cross-phase back-writing from concrete runtime evidence.

#### UI API design and layout

- [`compose-component-design`](skills/compose-component-design/SKILL.md) — design caller-placeable Compose APIs whose variable visual regions are caller-provided slots.
- [`compose-animations`](skills/compose-animations/SKILL.md) — choose Compose animation APIs for visibility, value targets, coordinated transitions, and content swaps; align with official quick guide and decision tree.
- [`compose-focus-navigation`](skills/compose-focus-navigation/SKILL.md) — design and test keyboard, TV, D-pad, and focus-first Compose navigation behavior.

#### Testing

- [`compose-ui-testing-patterns`](skills/compose-ui-testing-patterns/SKILL.md) — choose between plain UI tests, semantics assertions, key/focus tests, interaction state tests with MutableInteractionSource, screenshot tests, and integration tests.

### Kotlin

- [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) — review coroutine, raw `Thread`, and `Executor` ownership, cancellation, Flow state/event modeling, sharing, replay, and one-shot delivery.
- [`kotlin-control-flow`](skills/kotlin-control-flow/SKILL.md) — write and review Kotlin branching with subject `when`, guard conditions, sealed exhaustiveness, smart casts, nullable branching, and early returns.
- [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) — choose function owners, semantic domain types, and Kotlin Multiplatform platform boundaries.

### Writing

- [`grounded-writing`](skills/grounded-writing/SKILL.md) — draft or review public developer documentation and other user-owned text using Chris Banes’s writing as the sole style model, including internal report reviews, while preserving evidence, format, review-only authority, and material-edit restraint.

### Workflows

- [`deliver-spec`](skills/deliver-spec/SKILL.md) — coordinate one approved spec or verified Project-controller handoff through solo or justified delegated implementation, a planning-free fast path for small well-understood tickets, CI-supplied full-suite evidence, concurrent external review with verified draft support or earlier PR readiness, reusable scope approval and evidence, independent integrated-candidate review and focused repair review, and PR shepherding; validate applicable live-qualification harness paths offline (isolated production stores for persistent paths, fake runtimes for asynchronous paths), then reconcile failures under bounded approved allowances.
- [`release-kotlin-library`](skills/release-kotlin-library/SKILL.md) — assess readiness, prepare, and verify Kotlin library releases; check the `gradle-maven-publish-plugin` prerequisite, reconcile changelogs and Metalava API snapshots, and follow repository checks and publication gates.
- [`gradle-run`](skills/gradle-run/SKILL.md) — run every agent-initiated Gradle command through a compact-output wrapper; recover run identity and bounded results by workflow ID with read-only status, and diagnose failures with optional investigation helpers.
- [`implement-with-subagents`](skills/implement-with-subagents/SKILL.md) — coordinate one or more worker-owned implementation tasks through dependency-aware dispatch, task acceptance, integration, and same-owner repair; dispatch independent ready work concurrently within actual capacity, reuse sequential worker slots, and support read-only orchestration review.
- [`to-plan`](skills/to-plan/SKILL.md) — create risk-scaled implementation plans that prove baselines from green CI, with observable slices and consequential boundary validation before dependent work, plus authenticated integration amendments and explicit consumer handoffs; decide separate PRs by coherent deliverable boundaries.
- [`run-github-project`](skills/run-github-project/SKILL.md) — set up, review, or operate a repository's GitHub Project workflow with current-column authorization, human-only Backlog promotion, and deterministic configured agent selection; deliver ordinary implementation tickets through `deliver-spec`, resolve run-level merging once at startup with a 60-second default-no question unless already specified, preflight offline live-qualification harness evidence, preserve asynchronous or persistent failures in durable checkpoints, prioritize active delivery through safe planner checkpoints, fast-path small Todo tickets straight into implementation, start external review while CI runs after local qualification, free slots while reviewed PRs wait on CI, record ticket-local pauses while independent work continues, and keep `drain` alive under a read-only Project Watcher instead of returning at the first human wait.
- [`shepherd`](skills/shepherd/SKILL.md) — autonomously poll open PRs and MRs, or follow a host PR monitor's wake events (such as Claude desktop's Auto-fix) instead of polling, triage review comments, and repair CI failures with affected checks and required integrated-candidate validation; start external review during CI while retaining final-head merge gates.

Workflows that delegate agents share the [subagent selection and handoff
reference](references/subagent-selection.md). Each workflow retains its own
delegation trigger, authority, and acceptance rules. Each consuming skill also
contains a copy of the reference for standalone installation.
`deliver-spec` bundles its own solo-or-delegated implementation procedure;
`implement-with-subagents` retains its explicitly delegated workflow.

Planning and drain admission distinguish behavioral prerequisites, shared-file
mutation conflicts and final integration requirements. Plan-proven independent
preparation needs explicit slice ownership, permitted paths/resources, focused
checks and a stop before shared edits; native blockers, accepted-plan ordering,
resource grants and final integrated validation still apply.

Edit `references/subagent-selection.md`, then run
`npm run references:sync` and include the updated bundled copies in the change.
Run `npm run build` to check that all copies are regular files matching their
sources; it fails on drift without rewriting files.

### Migration from pre-cluster skills

This is a breaking taxonomy change. Replace the removed entrypoints as follows:

| Removed skills | Replacement |
|---|---|
| `compose-state-authoring`, `compose-state-hoisting`, `compose-side-effects` | [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) |
| `compose-recomposition-performance`, `compose-stability-diagnostics`, `compose-state-deferred-reads` | [`compose-performance`](skills/compose-performance/SKILL.md) |
| `compose-modifier-and-layout-style`, `compose-slot-api-pattern` | [`compose-component-design`](skills/compose-component-design/SKILL.md) |
| `kotlin-coroutines-structured-concurrency`, `kotlin-flow-state-event-modeling` | [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) |
| `kotlin-functions`, `kotlin-types-value-class`, `kotlin-multiplatform-expect-actual` | [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) |

## Contributing

Skills live at `skills/<skill-name>/SKILL.md`, flat (no language nesting). The `name:` in the SKILL.md frontmatter must match the directory name.

Frontmatter is validated against [`skills.schema.json`](skills.schema.json), which
tracks the core [Agent Skills specification](https://agentskills.io/specification)
and permits `disable-model-invocation` for Claude Code compatibility.
`name` and `description` are required; portable optional fields are `license`,
`compatibility`, `metadata`, and `allowed-tools`. Explicit-only workflow skills
also mirror that policy in Codex's `agents/openai.yaml`.

### Releases

Release versions use CalVer: `YYYY.M.D`, `YYYY.M.D.N`, or `YYYY.M.D.NN`, without
zero-padded month or day values. For example, use `2026.6.17` for the first
release of the day and `2026.6.17.1` or `2026.6.17.01` for another release.
Single-digit daily release numbers are normalized to the padded form, so both
inputs produce `2026.6.17.01`.

Keep root `plugin.json`, `.claude-plugin/plugin.json`,
`.codex-plugin/plugin.json`, and new Git release tags on the same version.
Existing zero-padded tags from before this policy map to the non-padded manifest
version, so `2026.06.16` maps to `2026.6.16`. Only bump versions when publishing
an installable release.

To publish a release, run the **Release** workflow from GitHub Actions. Leave the
version input empty to choose today's next unused UTC version, or provide a specific
CalVer value, optionally with a one- or two-digit daily release number. Use the
dry-run option to validate without creating a commit, tag, or GitHub release.
Automatic numbering checks existing tags and releases: the first release uses
`YYYY.M.D`, then subsequent releases increment the highest daily suffix through
`.99`. Gaps are not reused. If all daily numbers are exhausted, the workflow
fails and requires an explicit version or a release on another day.

The same workflow also runs nightly (02:17 UTC). It releases only if `skills/`,
the plugin manifests, `.opencode/`, `.agents/`, `plugin.json`, or `package.json`
changed since the latest tag; otherwise it exits without a release.

Before pushing, lint skills (frontmatter schema + markdown):

```
npm install
npm run lint
```

This also runs on CI for all PRs.

For a taxonomy change, also run the durable
[cluster behavior evaluation](tests/cluster-behavior.md). It checks routing,
required references, safeguards, exceptions, and finish gates at the public
agent-facing seam.

Before publishing a release, manually run the advisory evaluations for the
changed skills. Select each affected suite; when shared evaluation machinery
changes, include every suite affected by that change. Preview each suite with
the intended filters and current subject and judge cost assumptions:

```sh
python3 evals/run.py plan \
  --suite <suite> \
  --skill <changed-skill> \
  --model <model> \
  --reasoning <effort> \
  --judge-model <judge-model> \
  --judge-reasoning <effort> \
  --subject-cost-per-call-usd <amount> \
  --judge-cost-per-call-usd <amount> \
  --json
```

Repeat `--skill` for each changed skill. Without a `--case` filter, omitting
`--skill` selects all non-calibration cases in the suite. Calibration cases are
excluded unless you pass their explicit IDs with `--case`. Inspect `case_ids` in
the JSON output and confirm the selected cases match the intended scope before
reviewing call counts and estimated cost. The `plan` preview includes only the
counts and cost for first attempts. Execution may automatically retry each
subject and judge call once. Before `--execute`, get explicit cost approval for
up to twice the previewed subject calls, judge calls, and estimated cost, to
cover one retry per call. Execute the matching `run` command manually. Report
invalid or inconclusive results. Before any rerun, preview again, inspect its
`case_ids`, review the calls and cost, and get new approval for the same retry
headroom. Results remain advisory, not release gates. See
[`evals/README.md`](evals/README.md) for suite selection and command options.

## Evaluating skills

The advisory evaluator tests concrete scenarios modelled on real-world coding
work, with expected outcomes and no-change controls. It compares no-skill,
forced-skill, and automatic-routing runs. **Baseline** and **automatic** use
the cases eligible for automatic activation; **restraint** checks that a skill
does not make an unnecessary change. Scorecards also compare subject-side tokens,
tool calls, completed turns, elapsed time, and total attempted work per
successful outcome. The
table reports the latest available result for each skill and correctness metric.
These scores were produced using
[`gpt-6-luna`](https://developers.openai.com/api/docs/models/gpt-6-luna)
with high reasoning, judged by
[`gpt-6-sol`](https://developers.openai.com/api/docs/models/gpt-6-sol) with
high reasoning. Results are model- and reasoning-specific; other configurations
may perform differently. The human audit queue remains open. These are not merge
or release gates. The subject and judge each run on Codex by default, or on
Claude Code with `--runtime claude` and `--judge-runtime claude`; Claude runs need
an `--output-dir` outside the repository. See
[`evals/README.md`](evals/README.md) for evaluation setup and reproducibility.

Skill-revision compatibility checks compare old and revised instructions within
each model, separately from these benchmark scores. See the
[workflow compatibility record](evals/artifacts/2026-09-06-workflow-model-compatibility.md)
for Astra and 5.6 coverage and its current evidence limits.

| Skill | Baseline | Automatic | Restraint |
| --- | ---: | ---: | ---: |
| [`compose-animations`](skills/compose-animations/SKILL.md) | 75.0% | 100.0% | 100.0% |
| [`compose-component-design`](skills/compose-component-design/SKILL.md) | 86.7% | 100.0% | 100.0% |
| [`compose-focus-navigation`](skills/compose-focus-navigation/SKILL.md) | 33.3% | 100.0% | 100.0% |
| [`compose-performance`](skills/compose-performance/SKILL.md) | 83.3% | 100.0% | 100.0% |
| [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) | 83.3% | 100.0% | 100.0% |
| [`compose-ui-testing-patterns`](skills/compose-ui-testing-patterns/SKILL.md) | 55.6% | 100.0% | 100.0% |
| [`gradle-run`](skills/gradle-run/SKILL.md) | 41.7% | 100.0% | 100.0% |
| [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) | 58.3% | 100.0% | 100.0% |
| [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) | 44.4% | 100.0% | 100.0% |
| [`kotlin-control-flow`](skills/kotlin-control-flow/SKILL.md) | 33.3% | 100.0% | 100.0% |
| [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md) | 33.3% | 100.0% | 100.0% |
| [`deliver-spec`](skills/deliver-spec/SKILL.md) | — | — | — |
| [`grounded-writing`](skills/grounded-writing/SKILL.md) | 0.0% | 100.0% | 100.0% |
| [`implement-with-subagents`](skills/implement-with-subagents/SKILL.md) | — | — | 100.0% |
| [`release-kotlin-library`](skills/release-kotlin-library/SKILL.md) | 0.0% | 100.0% | 100.0% |
| [`run-github-project`](skills/run-github-project/SKILL.md) | — | — | 100.0% |
| [`shepherd`](skills/shepherd/SKILL.md) | — | — | 100.0% |
| [`to-plan`](skills/to-plan/SKILL.md) | — | — | 100.0% |

The `android-benchmark-comparison`, `compose-state-and-effects`,
`compose-ui-testing-patterns`, `gradle-run`, `grounded-writing`, `kotlin-api-design`,
`kotlin-concurrency-and-flow`, `kotlin-control-flow`, and
`release-kotlin-library` automatic cells, and the `to-plan` restraint cell,
use later focused evidence. Baseline and efficiency values use the complete
suite. The [improvement result record](evals/artifacts/2026-09-24-gpt6-improvement-results.md),
[targeted probe record](evals/artifacts/2026-09-24-gpt6-targeted-100-probes.md),
and [inline repair record](evals/artifacts/2026-09-24-gpt6-inline-repair-results.md)
give provenance and remaining failures.

### Skill efficiency

Values are per-run medians, baseline → automatic, followed by the automatic
percentage change. These subject-only measurements use the latest complete,
same-run evidence available for each suite and include failed runs and negative
controls. Baseline-to-automatic efficiency comparisons use only cases eligible
for automatic activation. Multi-skill scenarios contribute to every targeted
skill row. A turn is one completed top-level turn (a Codex `turn.completed` or a
Claude Code `result`); time remains environment-sensitive.
Run provenance and local scorecard paths are in the
[GPT-6 improvement result record](evals/artifacts/2026-09-24-gpt6-improvement-results.md).

| Skill | Tokens / run | Tool calls / run | Turns / run | Time / run |
| --- | ---: | ---: | ---: | ---: |
| [`compose-animations`](skills/compose-animations/SKILL.md) | 58.6k → 85.4k (+46%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 28.2s → 34.8s (+23%) |
| [`compose-component-design`](skills/compose-component-design/SKILL.md) | 48.7k → 73.4k (+51%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 24.4s → 29.5s (+21%) |
| [`compose-focus-navigation`](skills/compose-focus-navigation/SKILL.md) | 59.1k → 84.9k (+44%) | 5 → 5 (+0%) | 1 → 1 (+0%) | 28.9s → 36.6s (+27%) |
| [`compose-performance`](skills/compose-performance/SKILL.md) | 49.1k → 85.0k (+73%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 26.4s → 31.8s (+21%) |
| [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) | 60.1k → 89.0k (+48%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 28.2s → 37.8s (+34%) |
| [`compose-ui-testing-patterns`](skills/compose-ui-testing-patterns/SKILL.md) | 60.2k → 84.2k (+40%) | 5.5 → 5 (-9%) | 1 → 1 (+0%) | 27.1s → 26.8s (-1%) |
| [`gradle-run`](skills/gradle-run/SKILL.md) | 59.8k → 104.3k (+75%) | 4 → 7 (+75%) | 1 → 1 (+0%) | 27.6s → 45.3s (+64%) |
| [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) | 60.1k → 83.9k (+40%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 32.8s → 38.8s (+18%) |
| [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) | 60.0k → 74.0k (+23%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 26.5s → 28.0s (+6%) |
| [`kotlin-control-flow`](skills/kotlin-control-flow/SKILL.md) | 60.9k → 83.9k (+38%) | 5 → 5 (+0%) | 1 → 1 (+0%) | 26.2s → 38.5s (+47%) |
| [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md) | 46.9k → 53.8k (+15%) | 3 → 3 (+0%) | 1 → 1 (+0%) | 23.2s → 35.0s (+51%) |
| [`grounded-writing`](skills/grounded-writing/SKILL.md) | 35.6k → 56.9k (+60%) | 2 → 4 (+100%) | 1 → 1 (+0%) | 23.2s → 22.1s (-5%) |
| [`release-kotlin-library`](skills/release-kotlin-library/SKILL.md) | 70.2k → 88.3k (+26%) | 4 → 5 (+25%) | 1 → 1 (+0%) | 42.1s → 43.9s (+4%) |

### Claude Code

Claude Code: `claude-haiku-5-5` with high effort, judged by `claude-sonnet-5-5` with high
effort on Claude Code. The metrics are defined as above, and `—` marks a cell with no
recorded result. Run provenance and scorecards are in the
[Claude Code result record](evals/artifacts/2026-10-19-claude-haiku-5-5-sonnet-judge-results.md).

| Skill | Baseline | Automatic | Restraint |
| --- | ---: | ---: | ---: |
| [`compose-animations`](skills/compose-animations/SKILL.md) | 91.7% | 100.0% | 100.0% |
| [`compose-component-design`](skills/compose-component-design/SKILL.md) | 86.7% | 100.0% | 100.0% |
| [`compose-focus-navigation`](skills/compose-focus-navigation/SKILL.md) | 100.0% | 100.0% | 100.0% |
| [`compose-performance`](skills/compose-performance/SKILL.md) | 91.7% | 100.0% | 100.0% |
| [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) | 79.2% | 95.8% | 100.0% |
| [`compose-ui-testing-patterns`](skills/compose-ui-testing-patterns/SKILL.md) | 88.9% | 100.0% | 100.0% |
| [`gradle-run`](skills/gradle-run/SKILL.md) | 16.7% | 83.3% | 100.0% |
| [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) | 91.7% | 91.7% | 100.0% |
| [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) | 72.2% | 100.0% | 100.0% |
| [`kotlin-control-flow`](skills/kotlin-control-flow/SKILL.md) | 61.1% | 94.4% | 100.0% |
| [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md) | 33.3% | 100.0% | 100.0% |
| [`deliver-spec`](skills/deliver-spec/SKILL.md) | — | — | 100.0% |
| [`grounded-writing`](skills/grounded-writing/SKILL.md) | 0.0% | 100.0% | 100.0% |
| [`implement-with-subagents`](skills/implement-with-subagents/SKILL.md) | — | — | 100.0% |
| [`release-kotlin-library`](skills/release-kotlin-library/SKILL.md) | 16.7% | 100.0% | 100.0% |
| [`run-github-project`](skills/run-github-project/SKILL.md) | — | — | 100.0% |
| [`shepherd`](skills/shepherd/SKILL.md) | — | — | 100.0% |
| [`to-plan`](skills/to-plan/SKILL.md) | 16.7% | 100.0% | 100.0% |

Per-run medians, baseline → automatic, for the same configuration:

| Skill | Tokens / run | Tool calls / run | Turns / run | Time / run |
| --- | ---: | ---: | ---: | ---: |
| [`compose-animations`](skills/compose-animations/SKILL.md) | 51.8k → 85.9k (+66%) | 6 → 8 (+33%) | 1 → 1 (+0%) | 47.7s → 46.4s (-3%) |
| [`compose-component-design`](skills/compose-component-design/SKILL.md) | 52.2k → 72.8k (+40%) | 6 → 6 (+0%) | 1 → 1 (+0%) | 30.2s → 35.2s (+16%) |
| [`compose-focus-navigation`](skills/compose-focus-navigation/SKILL.md) | 62.2k → 88.9k (+43%) | 7.5 → 8 (+7%) | 1 → 1 (+0%) | 44.6s → 56.9s (+27%) |
| [`compose-performance`](skills/compose-performance/SKILL.md) | 94.3k → 103.8k (+10%) | 8 → 9 (+12%) | 1 → 1 (+0%) | 45.0s → 51.3s (+14%) |
| [`compose-state-and-effects`](skills/compose-state-and-effects/SKILL.md) | 72.4k → 118.4k (+64%) | 7 → 9 (+29%) | 1 → 1 (+0%) | 39.4s → 51.3s (+30%) |
| [`compose-ui-testing-patterns`](skills/compose-ui-testing-patterns/SKILL.md) | 87.0k → 88.3k (+2%) | 10 → 8 (-20%) | 1 → 1 (+0%) | 39.3s → 44.7s (+13%) |
| [`gradle-run`](skills/gradle-run/SKILL.md) | 75.6k → 105.6k (+40%) | 6 → 5 (-17%) | 1 → 1 (+0%) | 36.3s → 29.9s (-18%) |
| [`kotlin-api-design`](skills/kotlin-api-design/SKILL.md) | 96.7k → 101.8k (+5%) | 8 → 9 (+12%) | 1 → 1 (+0%) | 45.4s → 53.5s (+18%) |
| [`kotlin-concurrency-and-flow`](skills/kotlin-concurrency-and-flow/SKILL.md) | 111.2k → 145.2k (+31%) | 9.5 → 10 (+5%) | 1 → 1 (+0%) | 47.1s → 50.4s (+7%) |
| [`kotlin-control-flow`](skills/kotlin-control-flow/SKILL.md) | 119.5k → 138.1k (+16%) | 10 → 10 (+0%) | 1 → 1 (+0%) | 41.3s → 44.1s (+7%) |
| [`android-benchmark-comparison`](skills/android-benchmark-comparison/SKILL.md) | 37.7k → 66.3k (+76%) | 3 → 3 (+0%) | 1 → 1 (+0%) | 22.0s → 22.6s (+3%) |
| [`grounded-writing`](skills/grounded-writing/SKILL.md) | 46.6k → 73.6k (+58%) | 3 → 5 (+67%) | 1 → 1 (+0%) | 19.2s → 28.3s (+47%) |
| [`release-kotlin-library`](skills/release-kotlin-library/SKILL.md) | 55.5k → 60.7k (+9%) | 5 → 3 (-40%) | 1 → 1 (+0%) | 43.6s → 54.9s (+26%) |
| [`to-plan`](skills/to-plan/SKILL.md) | 22.9k → 53.4k (+133%) | 2 → 4 (+100%) | 1 → 1 (+0%) | 9.4s → 21.8s (+132%) |

## License

[Apache 2.0](LICENSE)