Evidence Formatter Skill

Formatting raw technology evidence into structured citations involves grouping technical signals and job-posting mentions with transparent reasoning, deduplicating redundant entries, and generating source attribution with URLs and timestamps for human review.

Evidence deduplication merges highly similar or redundant technology signals while preserving the strongest signal and maintaining original source attribution. This ensures structured, citation-ready documentation avoids repetitive entries.

Generating confidence-scored citations for multiple detected technologies requires inputting technologies_with_confidence data. The formatting process groups evidence by type, adds timestamps and URLs, and documents confidence limitations across all entries.

Documenting limitations for weak or insufficient technical signals involves adding caveats and recommendations directly into the formatted evidence output. This flags job-posting-only or low-quality evidence for human review.

Structuring technical signals and job-posting mentions for report generation requires grouping correlated evidence by type, applying confidence scoring, and producing human-readable citations with source attribution and transparent reasoning.

Evidence formatting works with optional historical context alongside technical signals and job-posting mentions. It processes correlated technology evidence to produce deduplicated, citation-ready documentation with timestamps and URLs.