ai-swarm@1.0.0
CAUTION· install onlyPublisher claims: "Multi-agent AI coding swarm orchestration. Plan parallel tasks, spawn Claude/Codex/Gemini agents in tmux sessions with git worktrees, auto-review, auto-integ..." Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Nothing executable in the bundle ran; the trace covers the install step only.- Tested
- every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-16 and scored it 72/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Auto-update or dynamic fetch configuration; Prompting user to enter sensitive credentials; Auto-update or dynamic fetch configuration. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 0 s · 1 scenarios · 0 deviationsStatic findings
git commit -m "docs: auto-update ESR + persist worklog for $BASE_SESSION" 2>/dev/null || trueFix: Pin versions and sources; vendor anything fetched at install time.chmod +x ~/workspace/swarm/*.shFix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.if ! kill -0 "$pid" 2>/dev/null; thenFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell."Enter your API key"Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.1. Read `pending-notifications.txt` → send to human → clearFix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.| `pulse-check.sh` | Detect and kill stuck agents |Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.git commit -m "docs: auto-update ESR + persist integration history ($(date +%Y-%m-%d))" 2>/dev/null || trueFix: Pin versions and sources; vendor anything fetched at install time.chmod +x "$INTEG_WRAPPER"Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.chmod +x "$RUNNER_SCRIPT"Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.kill -9 "$pid" 2>/dev/nullFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.result = subprocess.run(["tmux", "has-session", "-t", session],Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.task["completedAt"] = int(subprocess.run(["date", "+%s"], capture_output=True, text=True).stdout.strip()) * 1000Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.pr_result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.if (( running )) && kill -0 "$pid" 2>/dev/null; thenFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.eor_result = subprocess.run(Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.# Still running after timeout — force killFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.chmod +x "$REVIEW_WRAPPER"Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.- **Build status:** <ran npx vite build / gradle build? pass/fail>Fix: Pin versions and sources; vendor anything fetched at install time.Cost to the agent
Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.
Known vulnerabilities
Vulnerability advisories (OSV) cover npm and PyPI packages; this registry has no advisory database.
What would make this Verified
8 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Prompting user to enter sensitive credentials≈ +9 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- Unclaimed publisher≈ +8 on the totalClaim linkbag: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Auto-update or dynamic fetch configuration≈ +4 on the totalPin versions and sources; vendor anything fetched at install time.
- File permission modification command≈ +2 on the totalRequest the narrowest permission that works; replace wildcards with the exact paths or hosts.
- Process kill command≈ +2 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- File type combined with upload action≈ +2 on the totalRemove the call, or declare the host in frontmatter and explain in the README what is sent and why.