The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documentation describes workflows that ingest untrusted data from external sources, which could contain malicious instructions designed to influence the agent.
Ingestion points: The agent is instructed to process files located in 01-data-intake-review/incoming/ and .eml files in 03-offline-communications-triage/eml/ (referenced in SKILL.md).
Boundary markers: Absent; the provided prompt templates do not include delimiters or specific instructions to the agent to disregard commands embedded within the data files.
Capability inventory: The skill involves reading local files and writing markdown reports (defined in SKILL.md).
Sanitization: Absent; no validation or content filtering of the ingested files is recommended in the templates.
[PERSISTENCE]: The skill documentation suggests setting up cron jobs to automate tasks. While intended for user productivity, this illustrates a method for establishing recurring execution on the host system.
[COMMAND_EXECUTION]: The skill provides bash script examples for automating logs and instructions for the user to use chmod +x to make scripts executable.

openclaw-executive-assistant-workflows