The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions require the agent to manage a local record store by performing file system operations. These include creating directories, reading, writing, and appending JSONL files at paths like ${CLAUDE_PLUGIN_DATA}/records/ and a user-configured mirror folder.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it is designed to ingest and analyze external data, including thread history, local files, and specifications to generate its inquiry rungs.
Ingestion points: SKILL.md specifies reading from "the thread, the files, the spec, what has already been said."
Boundary markers: The skill relies on a structured markdown output format defined in the "Output" section.
Capability inventory: The agent is instructed to perform file system writes and directory creation as part of its recording logic.
Sanitization: No specific sanitization or escaping of the ingested external content is mentioned in the instructions.
[EXTERNAL_DOWNLOADS]: The documentation provides installation instructions and links to resources hosted on the author's GitHub repository (github.com/synapseradio/ai-skills). These are documented as part of the standard setup and maintenance for the skill.

generate-questions