The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to download and install an external CLI binary using npx -y @mvanhorn/printing-press install roam --cli-only or go install github.com/mvanhorn/printing-press-library/.... These resources are hosted within the author's (mvanhorn) scope and repository.
[COMMAND_EXECUTION]: The skill extensively uses the Bash tool to execute roam-pp-cli commands. These commands include data retrieval (transcripts, chat history) and administrative mutations (SCIM user management, event cancellation).
[DATA_EXFILTRATION]: The skill documents a built-in feature --deliver webhook:<url> which allows the agent to POST command results directly to an arbitrary external URL. Additionally, the feedback command can transmit data to a remote endpoint if the ROAM_FEEDBACK_ENDPOINT environment variable is configured.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8):
Ingestion points: Data is ingested through commands like roam-pp-cli grep, decisions, and transcript-fanout which read external chat messages and meeting transcripts from the Roam HQ API into the agent's context.
Boundary markers: There are no instructions or delimiters defined to warn the agent to ignore instructions embedded within the processed transcript or chat data.
Capability inventory: The skill has access to the Bash tool and commands capable of data mutation (chat-post, scim-diff --apply, onair-event-cancel) and external transmission (--deliver webhook:<url>).
Sanitization: There is no evidence of sanitization or filtering of the external content before it is processed by the agent.

pp-roam