Announcing Beacon: Self-improving memory for coding agents
← All skills
Skill

bilibili-browse

skill:whiteguo233/openbiliclaw#browse

Browse Bilibili pages using agent-browser for visual exploration and DOM interaction.

Verdict
Suspicious

Risky behavior the analysis could not rule out.

Analyzed Oct 27, 2026, 23:35 UTC at c3f682b

Dispute security verdict

If you believe this verdict is incorrect, submit a dispute and we'll review it.

1
findings
medium
highest severity
1
threat classes
2,226
files scanned
17/17
checks satisfied
0
package advisories

Why this verdict

  • suspicious: finding skill:skills/browse:skills/browse/SKILL.md:34:SUPPLY

Security findings 1

Supply chain

mediumContextual riskmedium confidence

The skill's Requirements section instructs the agent to globally install the npm package `@anthropic/agent-browser` with no version pin, an unpinned remote-code prerequisite install the agent may execute before the user's task.

agent-browser must be installed: `npm install -g @anthropic/agent-browser`
vector:instruction
Counter-argument considered

Strongest case against: the install is open in a Requirements section and the declared description already names agent-browser, so nothing is concealed. Strongest case for: the carve-out covers only pinned official-registry installs, this command has no version pin and the agent is told to run it globally, and no shown unit declares a mutable install step, so the supply-chain risk stands as claimed.

Coverage gaps 3

Parts of the artifact the static analysis could not fully examine.

context:unanalyzable_binary binary content is only byte-scanned
docs/plans/2026-07-16-event-capture-completion-plan.md
context:unanalyzable_binary binary content is only byte-scanned
packaging/icon.icns
context:unanalyzable_binary binary content is only byte-scanned
packaging/icon.ico