The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data such as customer interview transcripts, support tickets, and analytics to populate the Opportunity Solution Tree.
Ingestion points: Instructions in SKILL.md and references/ost-anti-patterns.md describe pulling content from interviews, tickets, and behavioral analytics.
Boundary markers: The instructions do not define specific delimiters or warnings to prevent the agent from following instructions embedded within the processed research data.
Capability inventory: The skill primarily generates text-based trees and executes a local validation script scripts/ost_validator.py on a provided JSON input.
Sanitization: There are no explicit filtering or sanitization steps mentioned for the ingested customer evidence.
[COMMAND_EXECUTION]: The skill provides instructions to run a local Python script scripts/ost_validator.py to audit the structure and quality of the generated tree.
Evidence: SKILL.md contains a bash snippet for executing the script using python3 with arguments for input files and output format.
Analysis: The provided source code for scripts/ost_validator.py was audited and found to contain only standard library imports and functional validation logic (JSON parsing, date checking, and string matching). It does not perform network operations, subprocess calls, or dynamic code execution.

opportunity-solution-tree