The Agent Skills Directory

[SAFE]: The skill reads and writes to specific markdown files (wiki/*.md and ~/.talos/profile.md) to manage project memory. These actions are transparent and align with the skill's stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external markdown files and uses it to provide context for the agent. This is an inherent attack surface for systems that persist and retrieve user-generated or project-related content.
Ingestion points: ~/.talos/profile.md, wiki/hot.md, wiki/INDEX.md, and individual files within wiki/.
Boundary markers: The skill does not define specific delimiters or warnings to the agent to ignore instructions embedded within the memory files.
Capability inventory: The skill has permissions to read and write files within the project directory and the ~/.talos/ directory.
Sanitization: No sanitization or validation of the markdown content is specified before it is processed by the agent.

memorizer