The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of the papr CLI to manage a local RSS database, perform full-text searches, and fetch content from the network. This include destructive commands such as papr unsubscribe <id> --yes and papr admin cleanup <days> --yes. Additionally, the skill's instructions for papr newsletter add involve passing a --password argument on the command line; while no secrets are hardcoded in the skill, this pattern is a security anti-pattern as it may expose credentials in shell history or process listings.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because its primary purpose is to ingest and summarize content from untrusted external sources (RSS feeds and newsletters). Maliciously crafted articles could attempt to manipulate the agent's behavior.
Ingestion points: Untrusted data enters the agent context through papr read, papr search, and papr list (SKILL.md).
Boundary markers: The instructions do not define any delimiters or provide 'ignore instructions' warnings to the agent when it processes external article content.
Capability inventory: The skill provides access to network operations (papr refresh), database modifications (papr settings set), and administrative data deletion (papr unsubscribe).
Sanitization: There is no evidence of sanitization or escaping of the feed content before it is processed by the agent.

papr-rss