The Agent Skills Directory

[DYNAMIC_EXECUTION]: The script scripts/ensure_publication_tooling.py performs dynamic imports as part of its environment verification process.\n
Evidence: The script uses importlib.import_module() to check for the presence of the pubfig and pubtab packages in the Python environment before attempting to use them.\n- [COMMAND_EXECUTION]: The skill uses shell commands for environment management and to perform data conversion tasks.\n
Evidence: The helper script scripts/ensure_publication_tooling.py utilizes subprocess.run to execute package managers like pip or uv for installing dependencies.\n
Evidence: Skill instructions in SKILL.md direct the agent to execute the pubtab CLI tool to convert Excel spreadsheets into LaTeX tables.\n- [EXTERNAL_DOWNLOADS]: The skill is configured to automatically download and install required Python libraries if they are not available in the local environment.\n
Evidence: The SKILL.md workflow and the ensure_publication_tooling.py script are designed to install the pubfig and pubtab packages from public registries to ensure the workflow is runnable.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external data sources which represents a potential attack surface for indirect prompt injection.\n
Ingestion points: According to SKILL.md, the skill ingests user-provided research results, Excel workbooks, LaTeX snippets, and raw data files.\n
Boundary markers: Absent; there are no specific delimiters or warnings provided to the agent to prevent it from following instructions that might be embedded within the research data.\n
Capability inventory: The skill has permissions to write files to the filesystem (e.g., PDF, PNG, SVG outputs) and execute CLI commands via the pubtab tool.\n
Sanitization: Absent; the skill relies on the internal parsing logic of the pubfig and pubtab libraries without performing explicit sanitization of the input content.

publication-chart-skill