Is Compose Skill safe?
- Sensitive path / secret-location access
- Python shell/command execution
- Python filesystem read
What to do: The Kotlin code was not checked for shell, network or file access, so this result is incomplete. Review that code before installing.
Compose Skill is an agent skill analyzed by SkillTotal's deterministic static scanner. The scan found no malicious indicators, though 1 risky construct is reported for review. Its Kotlin code was not checked for shell, network or file access, so this result is partial. It can: filesystem read, filesystem write and shell execution — capabilities are what the code can do, not a verdict on intent. Risk score 20/100 (low).
compose-skill
Get notified when this component changes
Change alerts aren't live yet. Leave your email and we'll let you know when they launch. We use it only for this and delete it on request.
Automated static-analysis result. It can contain false positives and false negatives, and is not a claim about the intent of Compose Skill's authors. Report a false positive.
Behavioral traits
How this component maps to the CSA agentic threat model. Descriptive — it never affects the risk score.
Findings (4)
A credential location (e.g. ~/.ssh, id_rsa) is passed to a file, process, or network call.
auth = os.path.expanduser("~/.codex/auth.json")Why it matters: This reads or ships a secret location, not merely mentions it — a common precursor to theft.
Fix: Verify why the component accesses credential locations; reading these is a common precursor to secret exfiltration.
The component can run operating-system commands or spawn processes.
tar = subprocess.run(["git", "-C", REPO, "archive", ref, src_rel],
capture_output=True, check=True).stdoutsubprocess.run(["tar", "-x", "-C", dest], input=tar, check=True)
p = subprocess.run(
["codex", "exec", "-s", sandbox, "--ephemeral", "--json",
"--skip-git-repo-check", "-C", workdir, "-o", out_md, prompt],
env=env, stdin=subprocess.DEVNULL, capture_output=True, text=T …Why it matters: Powerful and often legitimate — confirm the commands aren't built from untrusted input.
Fix: Confirm the command and its arguments are fully controlled and not derived from untrusted input; avoid shell=True.
The component reads files from disk.
with open(p, "rb") as fh:
cases = json.load(open(os.path.join(HERE, a.cases_file)))["cases"]
base_doc = json.load(open(a.baseline))
cand_doc = json.load(open(a.candidate))
s = open(SKILL, encoding="utf-8").read()
for n, line in enumerate(open(FIXTURES, encoding="utf-8"), 1):
Why it matters: Usually legitimate, but worth confirming it can't be steered into reading sensitive files.
Fix: Confirm which files are read and that paths cannot be influenced by untrusted input to reach sensitive locations.
The component writes or deletes files on disk.
shutil.copytree(os.path.join(REPO, src_rel), os.path.join(dest, "compose-expert"),
ignore=shutil.ignore_patterns(".DS_Store"))shutil.rmtree(os.path.join(dest, "skills"))
open(os.path.join(home, "config.toml"), "w").close()
shutil.copytree(target, dst, symlinks=True, ignore=shutil.ignore_patterns(
".gradle", "build", ".kotlin", ".git", ".idea", ".composeproof", "*.iml"))with open(os.path.join(trace_dir, name + ".jsonl"), "w") as fh:
shutil.rmtree(work, ignore_errors=True)
shutil.rmtree(base, ignore_errors=True)
json.dump(doc, open(out, "w"), indent=2)
Why it matters: Usually legitimate, but worth confirming the paths can't be controlled by untrusted input.
Fix: Confirm which files are written/deleted and that paths cannot be influenced by untrusted input.
How attackers abuse these capabilities
Interactive labs on the attack class behind the rules above. They show the technique, not anything found in Compose Skill.
Maintain Compose Skill?
The free Claude Code plugin scans each npm, PyPI or MCP package with this report's open-source engine before the install command runs. It runs on your machine and needs no account.
Check your own component
Run the same evidence-backed scan on any MCP server, agent skill, or package.
Scan your own componentHow we determine this: deterministic static analysis (regex + AST), evidence-anchored, no code execution. Methodology →