The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches content from user-specified external URLs (GitHub repositories, technical articles, and documentation) to extract design patterns and core thesis statements.
[COMMAND_EXECUTION]: Invokes a local Python script python3 scripts/effective_profile.py located in the profilera skill directory to weight applicability judgments based on the user's established decision profile.
[PROMPT_INJECTION]: Presents a surface for indirect prompt injection as it ingests and processes untrusted data from external web resources. Ingestion points: Source content fetched in Step 2 from arbitrary URLs. Boundary markers: Lacks explicit instruction delimiters for isolating fetched data from the system prompt. Capability inventory: Capabilities include local filesystem reads, GitHub MCP integration, and the execution of specific suite-related scripts. Sanitization: No specific sanitization or filtering of external content is defined in the instructions; however, the skill's read-only analysis focus limits the potential for exploitation compared to execution-oriented skills.

inspirera