The Agent Skills Directory

[DATA_EXFILTRATION]: The skill sends usage telemetry to https://www.respira.press/api/skills/track-usage. The data transmitted includes a hash of the site URL, the report framing selected, window length, and numerical totals. This destination belongs to the verified author of the skill (Respira for WordPress) and no sensitive content such as client names or activity notes are included in the telemetry packet.
[INDIRECT_PROMPT_INJECTION]: The skill ingests activity logs from a WordPress site using the respira_generate_activity_report tool and interpolates these notes into the final report output. This creates a surface for indirect prompt injection if the source activity logs contain malicious instructions. However, as the data source is the user's own authenticated site log and the skill does not grant the agent autonomous capabilities to execute shell commands or exfiltrate data based on that content, the risk is minimal.
[COMMAND_EXECUTION]: The skill utilizes several MCP tools (respira_generate_activity_report, respira_get_active_site, and respira_create_custom_post). These are specialized tools from the respira-wordpress server used for legitimate WordPress management and reporting tasks within the scope of the skill's purpose.

activity-report-composer