The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary workflow involves reading and processing various files within a user's research repository to understand the project and generate paper drafts. * Ingestion points: SKILL.md suggests using shell commands like find ., grep, and cat to explore repository structure, results, and documentation. * Boundary markers: The skill does not define specific delimiters or include instructions to disregard potential commands embedded within these untrusted files. * Capability inventory: The skill utilizes shell command execution and network capabilities through Python snippets and external search tools. * Sanitization: There is no documented process for sanitizing or escaping content read from the project files before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill involves fetching configuration and tooling from external locations. * The NeurIPS template's Makefile downloads conference style files from media.neurips.cc. * The instructions recommend installing an academic search tool from mcp.exa.ai. * Citation retrieval involves network requests to doi.org and Semantic Scholar.
[COMMAND_EXECUTION]: The skill provides instructions for the agent to execute shell commands to explore the research environment and manage LaTeX compilation, including ls, find, grep, pdflatex, and bibtex.

conference-paper-writing