The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing shell commands to perform deep searches across the local system. It uses standard utilities such as rg (ripgrep), grep, and ls, as well as a specialized CLI tool joelclaw (a vendor-specific resource) to query memory sources.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and synthesize content from external, untrusted sources including markdown notes in ~/Vault/, .svx files in ~/.brain, and previous session transcripts.
Ingestion points: Files located in ~/.brain and ~/Vault/, media metadata in /tmp/joelclaw-media/, and session data retrieved via joelclaw CLI.
Boundary markers: None identified. The instructions do not define delimiters or provide the agent with guidance to ignore potential instructions embedded within the retrieved notes or media tags.
Capability inventory: The skill possesses filesystem access (rg, grep, ls) and access to runtime telemetry and event logs via the joelclaw tool.
Sanitization: None. The workflow instructs the agent to synthesize findings and present them directly into the context, which could allow malicious instructions stored in notes to influence agent behavior.

recall