The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The pipeline ingests data from external sources including user-provided CSV files and crawled website content to identify business signals.
Ingestion points: Prospect data is read from CSV files in build_base.py, while website text is retrieved in run_sites.py and detect_signals_http.py.
Boundary markers: The skill does not implement explicit boundary markers or delimiters when interpolating external content into the final campaign text or generated documents.
Capability inventory: The agent has the capability to execute local scripts via subprocess, perform network requests, and write output files to the local file system.
Sanitization: While the scripts use regex patterns to extract specific technical markers (e.g., CMS types, job vacancy codes), raw text snippets and quotes from external websites are stored and eventually included in prospect outreach documents.
[COMMAND_EXECUTION]: The orchestrator script run_pipeline.py uses subprocess.call to execute the various stages of the data pipeline. These calls are restricted to the skill's own internal scripts and use the local Python interpreter.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to interact with the Apify API and directly fetches HTML content from company domains provided in the input prospect list to identify technology stacks and business signals.

prospect-list-run