Aggregate experiment analysis reports and execution logs into dated narrative digests.
npx skills add https://github.com/wanghao9610/STAR --skill star-expt-digest
Please help me install this Agent Skill. Skill: star-expt-digest Source: https://github.com/wanghao9610/STAR/tree/main/.cursor/skills/star-expt-digest Command: npx skills add https://github.com/wanghao9610/STAR --skill star-expt-digest
This Skill solves the difficulty of tracking research progress across multiple experiments, providing a narrative summary of what has changed, what was learned, and where the work stands.
Run the star-expt-digest skill to generate a progress report covering all experiments from the last seven days.
High-intent search queries and answers about installing and using this skill.
To generate a research progress report from experiment logs, use a tool that aggregates analysis reports and execution logs into a dated narrative digest. It automatically tracks changes in experiment verdicts and plan-level findings across defined time windows.
The best way to track changes in experiment verdicts is to summarize research experiment progress by aggregating logs into a digest. This identifies shifts in outcomes, such as refuted claims, ensuring no critical findings are overlooked before meetings.
A periodic experiment digest handles verified and provisional evidence by maintaining a strict separation between the two tiers. It provides structured provenance for all reported metrics while tracking changes across plan families.
Yes, you can aggregate experiment execution logs over a custom time window. The summarization process tracks changes in experiment outcomes and plan-level findings specifically across defined time windows or plan families.
Yes, you need structured provenance for all metrics in a research progress report to maintain a strict separation between verified and provisional evidence tiers. This ensures the reliability of reported experiment verdicts and findings.
Your experiment progress summary may not reflect the latest plan-level findings if the defined time window excludes recent execution logs. The digest generation strictly aggregates analysis reports based on the specified time parameters.
danielmiesslerStream PAI multi-agent activity to a WebSocket-enabled dashboard.
FixPortalConsolidate adversarial-review findings into a dated cross-repo digest.
kirvinExtract structured details from bug reports and create beads issues.
GiorgioRicciardielloRun parameter sweeps over computational scripts and generate summary reports.
