The Agent Skills Directory

[SAFE]: The skill defines a procedural workflow for managing an experimental journal. It involves reading and writing to specific local files (experiments/journal.md and experiments/lessons.md) within the project root. The skill does not involve network operations, execution of remote code, or access to sensitive system credentials.
[INDIRECT_PROMPT_INJECTION]: The skill establishes a data ingestion surface where the agent processes previously logged experimental data.
Ingestion points: The agent is instructed to read experiments/journal.md and experiments/lessons.md to review history and distill lessons (documented in SKILL.md).
Boundary markers: The skill does not define specific delimiters or warnings to ignore instructions that might be embedded in the journal logs.
Capability inventory: The skill's scope is limited to local file system reads and writes; it does not request network access, command execution, or elevated privileges.
Sanitization: The skill does not specify any sanitization, filtering, or validation of the journal data before the agent uses it for reasoning.

ml-experiment