The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill facilitates the installation of the Nextflow engine and test framework by piping remote scripts to a shell.
Evidence: Includes commands such as curl -s https://get.nextflow.io | bash in SKILL.md and curl -fsSL https://get.nf-test.com | bash in references/testing.md for tool setup.
[EXTERNAL_DOWNLOADS]: Fetches configuration files and software components from the official Nextflow and nf-core project domains.
Evidence: Downloads the main Nextflow launcher and nf-core pipeline assets from nextflow.io and nf-co.re as part of the documented setup and execution workflows.
[COMMAND_EXECUTION]: The skill is designed to manage and execute shell-based bioinformatics pipelines.
Evidence: Provides extensive documentation on writing script: blocks within Nextflow processes to execute tools like samtools, bwa, and fastqc (documented across all reference files).
[PRIVILEGE_ESCALATION]: Recommends the use of administrative privileges during the initial setup of the tool environment.
Evidence: Recommends running sudo mv nextflow /usr/local/bin/ to place the installed binary on the system path in the Setup section of SKILL.md.
[INDIRECT_PROMPT_INJECTION]: The skill provides the capability to ingest external data which influences the construction of shell commands.
Ingestion points: Processes external CSV samplesheets and configuration parameters via splitCsv and channel factories in SKILL.md, references/language.md, and references/running-pipelines.md.
Boundary markers: Instructional code examples do not explicitly define delimiters or "ignore instructions" markers for data interpolated into shell scripts.
Capability inventory: The agent can execute arbitrary shell scripts through Nextflow processes, write to the file system via publishDir, and interact with cloud APIs via configured executors.
Sanitization: No input validation or escaping for user-supplied data in samplesheets is implemented in the provided examples.
[DYNAMIC_EXECUTION]: Includes language features for capturing command output at runtime.
Evidence: Utilizes the eval() function in process output definitions to dynamically capture software version strings, as shown in references/language.md and references/developing.md.

nextflow