The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The script scripts/processon-diagram.mjs contains a hardcoded Pre-Shared Key (PSK) processon_mcp_psk_2026 used to generate an authentication code via AES-128-CBC encryption. While this appears to be a client-side key for the integration flow rather than a user-specific secret, hardcoding keys is a poor security practice.
[COMMAND_EXECUTION]: The skill instructions in references/diagram.md, references/document-mindmap.md, and references/mindmap.md direct the AI agent to execute local Node.js scripts (scripts/processon-diagram.mjs and scripts/processon-mindmap.mjs) to perform its primary functions. These scripts handle network communication and local state management.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it is designed to ingest and process untrusted data from various sources.
Ingestion points: The skill reads content from local files, attachments, web pages, meeting notes, and reports as specified in references/document-mindmap.md.
Boundary markers: The instructions do not define explicit delimiters or 'ignore embedded instructions' warnings for the data being processed.
Capability inventory: The skill can perform network requests to smart.processon.com and write to the local file system (e.g., ~/.processon-unified/ and os.tmpdir()).
Sanitization: There is no evidence of sanitization or filtering of the external content before it is processed or passed to the transformation scripts.

draw-processon