atlas-loop@0.1.15
CAUTION· install onlyPublisher claims: "Atlas Loop —— 通用的递归学习地图 Agent Skill:把任何陌生主题拆解为独立离线的交互式 HTML 地图,并为每个节点配上深入探究与递归子图提示。" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Bundled scripts ran with no arguments; no agent task scenario for this bundle.- Tested
- 8 of 71 bundled scripts run with no arguments (the rest were not run: per-bundle cap), with decoy credentials in the environment · every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
What changed since 0.1.12
vet-skills diff atlas-loop 0.1.12 0.1.15In the sandbox, this version reached for nothing 0.1.12 did not. Compared from the two recorded runs, not from source.
- No longer contacted
- registry.npmjs.org
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; bundled scripts ran with no arguments; no agent task scenario for this bundle) on 2026-11-22 and scored it 77/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Access to sensitive file paths; Access to sensitive file paths. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
…break catch close continue do d|0 else elseif for goto halt hang h|0 if job j|0 kill k|0 lock l|0 merge new open quit q|0 read r|0 return set s|0 tcommit throw troll…Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.node <skill_dir>/scripts/build_atlas.mjs --data /tmp/atlas-loop-<ts>/graph-data.json --out <output_html>Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.npx @liduchuan/atlas-loop build --data <graph-data.json> [--out <output.html>]Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.npx @liduchuan/atlas-loop install [--dir <skills-dir>]Fix: Pin versions and sources; vendor anything fetched at install time.…410-\u042F\u0430-\u044F\u0451\u0401_0-9]+",_="\u0434\u0430\u043B\u0435\u0435 "+"\u0432\u043E\u0437\u0432\u0440\u0430\u0442 \u0432\u044B\u0437\u0432\u0430\u0442\u044C\u0438\u0441\u043A\u043B\u044E\u0447\…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.npx @liduchuan/atlas-loop verify --html <output.html>Fix: Pin versions and sources; vendor anything fetched at install time.npx @liduchuan/atlas-loop dev [--port 5179] [--data <graph-data.json>]Fix: Pin versions and sources; vendor anything fetched at install time.…"and \u0438 else \u0438\u043D\u0430\u0447\u0435 endexcept endfinally endforeach \u043A\u043E\u043D\u0435\u0446\u0432\u0441\u0435 endif \u043A\u043E\u043D\u0435\u0446\u0435\u0441\u043B\u0438 endwhile \u043A\u0…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.return fs.readFileSync(tmpOut, 'utf8').replace(/import\.meta\.env/g, '({MODE:"development"})');Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.console.error('用法:node build_atlas.mjs --data <graph-data.json> [--out <output.html>]');Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.return res.outputFiles[0].text.replace(/import\.meta\.env/g, '({MODE:"development"})');Fix: Stop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.npx @liduchuan/atlas-loop link --parent <parent.html> --child <child.html> (--node <id|title> | --stream <id|title>)Fix: Pin versions and sources; vendor anything fetched at install time.npx @liduchuan/atlas-loop build --data <graph-data.json> [--out <output.html>]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
Not checked against OSV yet; the weekly check has not reached this version.
What would make this Verified
3 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Unclaimed publisher≈ +8 on the totalClaim liduchuan: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Access to sensitive file paths≈ +4 on the totalStop enumerating or reading credential locations; take secrets from a declared environment variable the user sets on purpose.
- New publisher≈ +3 on the totalTime clears this on its own after 14 and 30 days. Claiming the handle (+50) is the fastest way past it.
- 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.
- npx invocation of non-well-known package≈ +2 on the totalPin versions and sources; vendor anything fetched at install time.