Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Baoyu Post To X
Security checks across static analysis, malware telemetry, and agentic risk
Overview
The skill is coherent for posting to X, but it uses saved Chrome login sessions, explicit anti-bot bypass techniques, and external EXTEND.md settings that can affect profile choice or auto-submit behavior.
Install only if you intentionally want an agent to control a logged-in X account through Chrome. Use a dedicated Chrome profile, inspect any EXTEND.md files, keep preview mode as the default, and require explicit confirmation before publishing.
Static analysis
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Dangerous exec
Critical
- Finding
- Shell command execution detected (child_process).
Env credential access
Critical
- Finding
- Environment variable access combined with network send.
VirusTotal
61/61 vendors flagged this skill as clean.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
What this means
Misuse could publish unwanted public content, trigger account restrictions, or violate platform rules because the tool is designed to look less like automation.
Why it was flagged
The skill can automate public posting and publishing while explicitly bypassing bot-detection controls, which is high-impact even though it is disclosed and preview is the default.
Skill content
Posts text, images, videos, and long-form articles to X via real Chrome browser (bypasses anti-bot detection).
Recommendation
Use only with explicit user approval for each post, avoid bulk or unsolicited posting, and consider a separate X account or Chrome profile.
What this means
Anyone or any agent workflow using this skill with your saved profile may be able to compose or publish from your logged-in X account.
Why it was flagged
The skill acts through a persistent local browser session and can use a selected Chrome profile, giving it delegated access to the user's X account.
Skill content
First run: log in to X manually (session saved) ... `--profile <dir>` | Custom Chrome profile
Recommendation
Use a dedicated Chrome profile with only the intended X account, do not point it at your everyday browser profile, and revoke or clear the profile if you no longer trust the skill.
What this means
A project file could change posting behavior or profile selection in ways the user did not intend, including pushing toward automatic submission.
Why it was flagged
A project-level or user-level EXTEND.md file can influence which browser profile is used and whether posting is automatic, but the artifacts do not show a strict schema, trust boundary, or approval check.
Skill content
Found | Read, parse, apply settings ... EXTEND.md Supports: Default Chrome profile | Auto-submit preference
Recommendation
Inspect any .baoyu-skills/baoyu-post-to-x/EXTEND.md file before use, disable auto-submit preferences, and require explicit confirmation before any --submit action.
What this means
The skill may require accessibility/automation permissions and can paste clipboard contents into a real application window.
Why it was flagged
The skill uses OS automation commands to send real paste keystrokes. This is disclosed and purpose-aligned for image/article pasting, but it can affect the active application if misdirected.
Skill content
const result = spawnSync('osascript', ['-e', script], { stdio: 'pipe' });Recommendation
Keep the Chrome window visible, verify the focused app before pasting, and avoid granting broad accessibility permissions unless needed.
What this means
The command may fetch or use tooling outside the reviewed skill files, depending on the local setup.
Why it was flagged
The documented execution path uses npx -y bun rather than a pinned local runtime. This is common for CLI workflows but means runtime resolution depends on the user's environment and package source.
Skill content
npx -y bun ${SKILL_DIR}/scripts/x-browser.ts "Hello!" --image ./photo.pngRecommendation
Install Bun from a trusted source, prefer a pinned/local runtime where possible, and verify the reviewed scripts before running them.