The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the tinyfish CLI to perform web searches and repository health checks. It also utilizes standard shell utilities such as cat, wait, and echo for managing temporary JSON results stored in /tmp/.
[EXTERNAL_DOWNLOADS]: The skill references and fetches data from established external platforms including GitHub and Reddit to evaluate software alternatives. These operations are part of the skill's primary function to aggregate public project metadata.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external web sources, which is a potential vector for indirect prompt injection. The skill mitigates this risk by providing specific instructions to the extraction agents to limit their scope and by advising that all external content be treated as untrusted for synthesis only.
Ingestion points: Web content retrieved from GitHub search, curated repository lists, and Reddit discussion threads (SKILL.md).
Boundary markers: The skill uses environment descriptions and specific output format requirements (JSON) for the extraction agents.
Capability inventory: Execution of tinyfish for web access and repository analysis; shell operations for file management.
Sanitization: Content is marked as untrusted and intended for synthesis by the LLM without further execution.

oss-alternatives