{ "schema_version": 2, "kind": "shortening-method", "format": "agent-skill", "id": "timeline", "name": "Timeline", "category": "Writing", "summary": "Condense a narrative into dated events without inventing chronology or causation.", "use_cases": [ "Project histories", "Event chronologies", "Case timelines" ], "word_count": 146, "url": "https://sho.rten.it/methods/timeline/", "instructions_url": "https://sho.rten.it/methods/timeline/SKILL.md", "skill_url": "https://sho.rten.it/methods/timeline/SKILL.md", "json_url": "https://sho.rten.it/methods/timeline/llms.txt", "plain_text_url": "https://sho.rten.it/methods/timeline/prompt.txt", "license": "MIT", "sources_url": "https://sho.rten.it/sources/#timeline", "skill_name": "timeline", "skill_description": "Condense a narrative into dated events without inventing chronology or causation. Use for Project histories, Event chronologies, Case timelines.", "agents_md_url": "https://sho.rten.it/methods/timeline/AGENTS.md", "sources": [], "instructions": "Turn the supplied narrative into a concise timeline of relevant events. Give each entry a date or time, a short event description, and a status when it is not complete. Include the responsible actor when needed to distinguish events.\n\nPreserve exact dates, time zones, date ranges, uncertain timing, and dependencies that affect interpretation. Order events only as precisely as the source permits. Group undated events separately or keep their explicit relative timing; do not assign a date by guesswork. Distinguish planned, scheduled, attempted, completed, and cancelled events. A later event does not by itself prove that an earlier event caused it.\n\nMerge duplicate accounts of the same event, retaining material disagreements and their sources. Remove narrative transitions and repeated background. Keep a supporting detail only when it explains a milestone, delay, or unresolved state. Return only the timeline and brief timing caveats needed to read it.", "example": { "context": "Sensor-trial chronology", "before": "The team approved the sensor trial on 2 June 2026. A shipment was scheduled to arrive on 9 June, but it actually arrived on 11 June. Installation was completed on 14 June. The notes mention a calibration problem sometime after installation and before the first test, but do not give a date. The first test ran on 18 June and failed its accuracy check. A second test is planned for 25 June, provided recalibration is complete. There is no recorded outcome for that second test yet.", "after": "- 2 June 2026: Trial approved.\n- 11 June: Shipment arrived; scheduled for 9 June.\n- 14 June: Installation completed.\n- After installation, before 18 June: Calibration problem noted; exact date unknown.\n- 18 June: First test failed the accuracy check.\n- 25 June: Second test planned, conditional on recalibration. No outcome recorded.", "must_preserve": [ "2026 dates and two-day delivery delay", "Calibration problem has only relative timing", "First test failed accuracy check", "Second test is planned and conditional, not completed" ], "omitted": [ "Cuts transitions and repeated references to the team and notes." ], "kind": "illustrative" } }