The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to set up the environment and run analysis.
Evidence: Instructions in SKILL.md specify running npm install and executing the Node.js sampler script using node.
[EXTERNAL_DOWNLOADS]: The skill installs third-party libraries from the NPM registry to enable WebSocket communications.
Evidence: npm install ws https-proxy-agent is used to prepare the environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes live, untrusted data from the Bluesky firehose, which could theoretically contain malicious instructions designed to influence the agent's behavior during analysis.
Ingestion points: The scripts/zeitgeist-sample.js script connects directly to the Bluesky Jetstream WebSocket (wss://jetstream1.us-east.bsky.network).
Boundary markers: None are employed when the script processes or when the agent analyzes the resulting JSON data.
Capability inventory: The skill uses node for data collection and instructions suggest using str_replace to generate an HTML artifact saved to /mnt/user-data/outputs/zeitgeist.html.
Sanitization: While the script tokenizes text for frequency analysis (filtering non-alphanumeric characters), it includes raw sample post text in the final data object without sanitization.

sampling-bluesky-zeitgeist