The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes completion data from tasks, tool calls, and phases, which serves as an entry point for untrusted data that could influence reporting or downstream analysis.
Ingestion points: Metric capture logic retrieves data from agent completions and tool call outputs in SKILL.md.
Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions for the data being logged.
Capability inventory: The skill performs file append operations to analytics.jsonl and generates report files in weekly_report.md.
Sanitization: The skill mandates PII scrubbing in telemetry and requires 'hard math validation' via Python scripts to prevent the LLM from hallucinating or being manipulated during report generation.

agent-analytics