Build chat applications with PubNub Chat SDK
97
72%
Does it follow best practices?
Impact
100%
1.14xAverage score across 20 eval scenarios
High
Do not use without reviewing
Security
1 high severity finding. You should review these findings carefully before considering using this skill.
The skill handles credentials insecurely by requiring the agent to include secret values verbatim in its generated output. This exposes credentials in the agent’s context and conversation history, creating a risk of data exfiltration.
Insecure credential handling detected (high risk: 1.00). The prompt shows API keys/authKey as string literals in the initialization example and requires the LLM to include SDK initialization with keys in outputs, which encourages embedding secret values verbatim (high exfiltration risk).
[REDACTED]
secret · 21 sites
Placeholder publishKey value '[REDACTED]' appears in 8 code examples across SKILL.md and chat-setup.md; the Output Format requires including SDK initialization in outputs, so an agent would emit real keys here.
Low
Low-risk findings.
1 low severity finding. Worth noting, but not necessarily harmful.
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to connect to PubNub channels and read user-generated messages (see SKILL.md core workflow "Connect to Channel" and references/chat-features.md use of channel.connect, getHistory, and public channels), which are untrusted third‑party content the agent would parse and could influence subsequent actions.
pubnub.com
domain · 9 sites
The scanner flags PubNub channel messages (via channel.connect, getHistory) as untrusted third-party content, but PubNub is the intended backend service this plugin integrates with - the messages are the application's own chat data, not arbitrary untrusted content from the open web.