The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Bash scripts (pubmed_eutils.sh) and Python scripts (parse_pubmed.py) located within the ~/.claude/skills/search-lit/references/ directory to interface with the PubMed E-utilities API.
[DYNAMIC_EXECUTION]: The skill uses python3 -c to execute an inline Python script for parsing MeSH terms from XML data retrieved from PubMed. The script logic is statically defined within the skill instructions and processes data from standard input.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests untrusted data from external sources including PubMed, Consensus, Scholar Gateway, and bioRxiv/medRxiv preprints.
Ingestion points: External data is retrieved via the PubMed E-utilities API and various Model Context Protocol (MCP) search tools (Consensus, Scholar Gateway, bioRxiv).
Boundary markers: The instructions do not specify the use of delimiters or warnings to ignore embedded instructions in the retrieved research data.
Capability inventory: The skill possesses capabilities for shell command execution (Bash), file writing (Write), and network access to literature databases.
Sanitization: There are no explicit sanitization or filtering steps documented for the content retrieved from external research databases before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch data from well-known and established academic services, including PubMed (NCBI), Consensus, Scholar Gateway, and the bioRxiv/medRxiv preprint servers. These operations are essential for the skill's primary purpose of literature discovery and gap analysis.

ma-scout