Agent skill

Audit Jmm

by ben-manes in ben-manes/caffeine

Java Memory Model audit of all VarHandle/volatile field access modes

Apache-2.0Auto-check passed

Install Audit Jmm

skills CLI
$ npx skills add ben-manes/caffeine --skill audit-jmm -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install ben-manes/caffeine audit-jmm --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/ben-manes/caffeine.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/audit-jmm .claude/skills/audit-jmm && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
audit-jmm
GitHub stars
18k
Token cost
~509 tokens
SKILL.md length
267 words
Files
1
Skills in repo
33
Repo updated
First seen
Licence
Apache-2.0

At a glance

Java Memory Model audit of all VarHandle/volatile field access modes

  • Works in 4 steps: State the field name, type, and declared… → List every read and write with: access… → For each (write, read) pair: state… → …
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Audit Jmm is an agent skill from ben-manes/caffeine. Java Memory Model audit of all VarHandle/volatile field access modes

Its SKILL.md is about 510 tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It works with Java. The repository describes itself as: A high performance caching library for Java. The licence is Apache-2.0.

Example prompts

  • “/audit-jmm”

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. State the field name, type, and declared access mode.
  2. List every read and write with: access mode, method/line, locks held.
  3. For each (write, read) pair: state whether happens-before is guaranteed.
  4. Identify any field where

What it can do on your machine

Read from SKILL.md and the folder at commit e972fb0. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Audit Jmm loads about 509 tokens when it runs. Until then it costs about 20 tokens; SKILL.md has 267 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~20
When it runs · the whole SKILL.md, loaded when a task matches
~509

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from ben-manes/caffeine at commit e972fb0, republished under its Apache-2.0 licence (© ben-manes). 267 words, ~509 tokens.

Download SKILL.mdSave it as .claude/skills/audit-jmm/SKILL.md (or your agent's skills folder).
name
audit-jmm
description
Java Memory Model audit of all VarHandle/volatile field access modes
context
fork
agent
auditor
disable-model-invocation
true

Perform a Java Memory Model audit of the cache.

For every field accessed via VarHandle or Unsafe, and for every field declared volatile or accessed under synchronization:

  1. State the field name, type, and declared access mode.
  2. List every read and write with: access mode, method/line, locks held.
  3. For each (write, read) pair: state whether happens-before is guaranteed. If it depends on access mode, verify BOTH sides use compatible modes.
  4. Identify any field where:
    • Plain/opaque write paired with plain/opaque read across threads with no intervening synchronization
    • Code relies on opaque providing ordering beyond coherence
    • Volatile read on one field but non-volatile on a correlated field

Specific areas to examine:

  • writeTime: encodes both timestamp and refresh-in-progress flag
  • accessTime: is opaque access sufficient for expiration?
  • key/value fields: do value reads provide visibility of the object's fields?
  • policyWeight vs weight: correlated but updated at different times

For each issue:

  • State the specific reordering or visibility failure
  • Construct a concrete 2-thread execution
  • Verify the execution is legal under the JMM (not just TSO)

Do not report issues that only affect performance. Do not report deliberately racy patterns with documented stale-read tolerance.

Platform focus: Pay specific attention to aarch64 (ARM) memory ordering. ARM's weaker-than-TSO model exposes reordering bugs invisible on x86. Historical bug (#1820): a weak/soft-value setValue published the new WeakValueReference with setRelease and then cleared the old reference. Without a storeStoreFence between them, aarch64 could make the clear() visible before the publication, so a reader saw a null value. Check for similar patterns: a release store followed by a store that a racing reader must not observe first.

© ben-manes, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .claude/skills/audit-jmm of ben-manes/caffeine.

Open the folder on GitHubat commit e972fb0

Compare with similar skills

Audit Jmm next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Audit Jmm compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Audit Jmm this skillben-manes/caffeine18k—~509Automated safety check: PassApache-2.0
Brainstormingxpinjection/test-driven-spring-boot11252 repos~2.6kAutomated safety check: PassMIT
Android API Diffgkd-kit/gkd43k—~796Automated safety check: PassGPL-3.0
Video Cover Imageitwanger/toBeBetterJavaer18k—~3.3kAutomated safety check: PassNone
Lancedb Update Lance Dependencylancedb/lancedb12k—~1.1kAutomated safety check: PassApache-2.0
Java SDK E2E Test with Replay Snapshotgithub/copilot-sdk11k—~1.8kAutomated safety check: PassMIT

Similar skills

  • Brainstorming

    xpinjection/test-driven-spring-boot

    You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior.

    112 GitHub starsUsed in 52 repos~2.6k tokens
    Agent WorkflowsAuto-check passed
  • Android API Diff

    gkd-kit/gkd

    Looks up Android framework Java and AIDL APIs across versions with the android-api-diff CLI: signatures, availability, source files and hidden-API access code.

    43k GitHub stars~796 tokensUpdated today
    MobileAuto-check passed
  • Video Cover Image

    itwanger/toBeBetterJavaer

    Generate matched 3:4, 16:9, and 4:3 short-video cover images from toBeBetterJavaer video scripts or AI/Java technical topics.

    18k GitHub stars~3.3k tokensUpdated today
    Media & CreativeAuto-check passed
  • Update LanceDB to a specific Lance release or tag. An agent skill from lancedb/lancedb.

    12k GitHub stars~1.1k tokensUpdated today
    DatabasesAuto-check passed
  • Official

    Creates a Java SDK end-to-end test for the Copilot SDK that runs against a recorded YAML snapshot through a replay proxy, so CI needs no real authentication.

    11k GitHub stars~1.8k tokensUpdated today
    Testing & QAAuto-check passed
  • Fory Release

    apache/fory

    Prepare an Apache Fory release candidate from a clean release branch, including the version bump, RC tag, JVM staging, ASF source artifacts, SVN upload, and vote email.

    4.6k GitHub stars~2.9k tokensUpdated today
    Auto-check passed

More from ben-manes/caffeine

All 33 skills in this repo
  • Runs controlled JMH experiments on the Caffeine cache to find shared contention and hot-path waste, then reviews correctness and returns a reviewable patch.

    18k GitHub stars~2.6k tokensUpdated yesterday
    Auto-check: notes
  • Git History Bug Audit

    ben-manes/caffeine

    Audits a module by walking its git history commit by commit, tracking unresolved issues forward, and reporting the ones that survive to HEAD as findings.

    18k GitHub stars~3.3k tokensUpdated yesterday
    Auto-check passed
  • Adversarial Codebase Audit

    ben-manes/caffeine

    Runs a hostile review of the Caffeine Java caching library with parallel subagents that get no design docs, then challenges and consolidates their findings.

    18k GitHub stars~1.9k tokensUpdated yesterday
    Auto-check: notes
  • Caffeine Performance Audit

    ben-manes/caffeine

    Audits the Caffeine cache source for hot-path costs such as allocations, contention and memory layout, reporting only findings tied to specific lines.

    18k GitHub stars~559 tokensUpdated yesterday
    Auto-check passed
  • Audit Sibling Divergence

    ben-manes/caffeine

    Compares code paths that should behave the same, such as sync and async cache methods, and requires a concrete scenario where the two observably disagree.

    18k GitHub stars~4.3k tokensUpdated yesterday
    Auto-check: notes
  • Climber Step Minimization

    ben-manes/caffeine

    Prices each step of the window climber algorithm by disabling it in turn, to find steps that no longer earn their keep and branches that no longer fire.

    18k GitHub stars~3k tokensUpdated yesterday
    Auto-check: notes

Works with

Questions about Audit Jmm

What does Audit Jmm do?

Java Memory Model audit of all VarHandle/volatile field access modes. Audit Jmm is an agent skill from ben-manes/caffeine.

How do I install Audit Jmm in Claude Code?

Run `npx skills add ben-manes/caffeine --skill audit-jmm -a claude-code`. Or copy the skill folder (.claude/skills/audit-jmm in ben-manes/caffeine) into .claude/skills/audit-jmm in your project. Claude Code loads it when a task matches its description.

How do I install Audit Jmm in Codex?

Run `npx skills add ben-manes/caffeine --skill audit-jmm -a codex`. Or copy the skill folder (.claude/skills/audit-jmm in ben-manes/caffeine) into .agents/skills/audit-jmm in your project. Codex loads it when a task matches its description.

Can I use Audit Jmm in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add ben-manes/caffeine --skill audit-jmm -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/audit-jmm, .gemini/skills/audit-jmm, .github/skills/audit-jmm and .opencode/skills/audit-jmm in your project.

What does Audit Jmm need to run?

SKILL.md names no scripts, command-line tools or credentials: Audit Jmm is instructions for the agent only.

Does Audit Jmm access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Audit Jmm safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Audit Jmm use?

Audit Jmm is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Audit Jmm use?

About 509 tokens (SKILL.md is roughly 2k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Audit Jmm?

Skills that share tags, products or a category with Audit Jmm: Brainstorming (xpinjection/test-driven-spring-boot, 112 stars), Android API Diff (gkd-kit/gkd, 43k stars), Video Cover Image (itwanger/toBeBetterJavaer, 18k stars) and Lancedb Update Lance Dependency (lancedb/lancedb, 12k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Audit Jmm?

ben-manes (a GitHub user) maintains it in ben-manes/caffeine, which has 17,882 GitHub stars. The repository holds 33 skills in this directory. The repository was last updated on October 9, 2026.

Source: ben-manes/caffeine on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.