The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from X (formerly Twitter) using the X.AI API to generate new content, creating a vulnerability to indirect prompt injection attacks.
Ingestion points: External social media content is retrieved from https://api.x.ai/v1/responses within the DRAFTS and REMIX branches.
Boundary markers: The instructions do not specify any delimiters or directives for the agent to treat external content as untrusted or to ignore instructions embedded in the fetched text.
Capability inventory: The skill possesses network access (curl), file system access (memory/ and soul/ directories), and the ability to communicate with the user via ./notify.
Sanitization: No sanitization or validation is performed on the content of the fetched tweets before it is used for content generation.
[COMMAND_EXECUTION]: The skill uses various shell utilities including curl, jq, sed, grep, and date to manage data processing, API interaction, and draft storage.
[EXTERNAL_DOWNLOADS]: The skill fetches content from the X.AI API to provide context and candidates for tweet generation, representing an external network dependency.

write-tweet