The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the joelclaw CLI tool to capture discoveries. It passes user-provided URLs and context strings as data payloads to shell commands.
[EXTERNAL_DOWNLOADS]: The skill triggers a background pipeline that performs automated downloads and repository cloning of external sources provided by the user via URLs.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external URLs (articles, repositories, and social media posts), creating a potential surface for indirect prompt injection attacks where malicious instructions could be embedded in the captured content.
Ingestion points: External URLs shared by the user in chat or auto-captured from specific Slack channels as described in SKILL.md.
Boundary markers: The skill instructions do not specify the use of delimiters or explicit instructions to ignore embedded commands within the captured data.
Capability inventory: The skill facilitates command execution via the joelclaw CLI, performs local file system writes to a Vault, and initiates network-based enrichment tasks.
Sanitization: The instructions do not describe any sanitization, validation, or filtering of the content retrieved from external sources before it is summarized or stored.

discovery