The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied novel content, creating a surface for indirect prompt injection where instructions embedded in the text could influence agent behavior.
Ingestion points: SKILL.md Phase 1 Step 3 reads novel content from user-provided paths or direct text inputs.
Boundary markers: Novel text is not wrapped in explicit delimiters or accompanied by instructions to ignore embedded commands.
Capability inventory: The skill has file system access to create and manage the project structure using scripts/tracking_commit.py.
Sanitization: scripts/tracking_commit.py performs strict schema validation and sanitizes file names to avoid illegal characters and reserved system names.
[COMMAND_EXECUTION]: The skill invokes a local Python utility script to manage project data.
Evidence: SKILL.md instructs the agent to execute scripts/tracking_commit.py using the host's Python interpreter for initializing and updating the project tracking state.

story-import