The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (scripts/init_project.py, scripts/migrate_project_to_ssot.py, scripts/validate_project_contract.py) using the Bash tool to automate project scaffolding, migration, and contract validation. These scripts are internal components of the skill's infrastructure.
[EXTERNAL_DOWNLOADS]: The skill documentation indicates a dependency on the pyzotero Python library for optional Zotero integration, which is a well-known service for researchers.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks as it reads and processes manuscript content and project metadata to generate status reports and checklists.
Ingestion points: Reads from project_state.json, SSOT.yaml, and various manuscript files (e.g., manuscript/index.qmd, sections/*.qmd).
Boundary markers: There are no explicit delimiters or specific instructions for the agent to treat ingested manuscript text as data rather than instructions.
Capability inventory: The skill possesses Bash (command execution), Write, and Edit capabilities across its various functions.
Sanitization: No explicit content sanitization or validation of the data ingested from manuscript files is described in the instruction files.

manage-project