The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted production call transcripts via mcp__cekura__call_logs_retrieve to identify failure modes and draft new scenario instructions. Malicious content within a transcript could theoretically attempt to influence the agent's logic during the clustering or drafting phase.
Ingestion points: Transcription data retrieved from mcp__cekura__call_logs_retrieve in the Step 1 and Step 3 workflows.
Boundary markers: No explicit delimiters or instructions to ignore embedded prompts within the transcripts are provided in the instructions.
Capability inventory: Access to mcp__cekura__scenarios_create, mcp__cekura__call_logs_create_scenarios, and file system tools (Write, Edit, Bash).
Sanitization: The skill does not explicitly sanitize or filter the transcript content before interpolating it into drafting logic or report generation.
[COMMAND_EXECUTION]: The skill uses Bash, Write, and Edit tools to generate and save a markdown report (failure_scenarios_<agent_id>.md) in the local working directory. This is a standard operational requirement for the skill's reporting functionality.
[SAFE]: External references to dashboard.cekura.ai and docs.cekura.ai are official vendor domains associated with the skill's author and are used for legitimate documentation and dashboard linking purposes.

cekura-generate-scenarios