The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from external sources which could contain malicious instructions.
Ingestion points: Processes markdown notes, journal reflections, and transcripts provided by the user (found in SKILL.md).
Boundary markers: Absent. The instructions do not specify the use of delimiters (like XML tags or triple backticks) or "ignore embedded instructions" warnings for the ingested content.
Capability inventory: The skill executes shell commands via uv run python and writes files to the local file system (found in SKILL.md and scripts/generate_principles.py).
Sanitization: Absent. There is no logic provided to sanitize or escape the content of the notes before the agent processes them into a JSON payload.
[COMMAND_EXECUTION]: The skill uses shell command execution to perform its primary functions.
Evidence: The Execution Protocol in SKILL.md instructs the agent to run uv run python to execute the bundled generate_principles.py script for rendering and validating reports. While the scripts are provided within the skill, this pattern utilizes the agent's ability to run local code.

generate-principles