The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection because it processes untrusted data from user-provided logo files (SVG/PNG) and external competitor websites.
Ingestion points: scripts/analyze-logo.py, scripts/inspect_svg.py, and scripts/prepare_logo_export.py analyze binary and text data from logo files. The research workflows (workflows/brand-research.md, workflows/competitor-profiling.md) ingest content from live web research.
Boundary markers: There are no explicit instructions to the agent to ignore or delimit embedded instructions within the processed files or research data.
Capability inventory: The skill can execute local Python scripts, perform network requests to well-known domains, and write files to the workspace.
Sanitization: No semantic sanitization is performed on text extracted from external sources before it is used by the agent for generating brand assets or strategies.
[SAFE]: The skill uses Python scripts to perform network lookups against official APIs and domains of well-known services (GitHub, npm, X, and IANA) to support its brand naming and domain checking functionality.
[SAFE]: Access to the X_BEARER_TOKEN environment variable in scripts/x-name-check.py follows standard security practices for API authentication and is used solely for requests to the official X API.

chrisai-branding