Analyze ARIS usage logs and propose evidence-backed SKILL.md patches.
npx skills add https://github.com/Shallow-W/llm-wiki --skill meta-optimize-shallow-w
Please help me install this Agent Skill. Skill: meta-optimize Source: https://github.com/Shallow-W/llm-wiki/tree/main/.claude/skills/meta-optimize Command: npx skills add https://github.com/Shallow-W/llm-wiki --skill meta-optimize-shallow-w
ARIS harness performance and user experience degrade over time when defaults, review prompts, and workflow parameters stop matching how people actually use the system.
.aris/meta/events.jsonl to identify which skills, parameters, and prompts correlate with overrides and failures.SKILL.md files and workflow defaults, with explicit log evidence and cross-model review gating..aris/meta/optimizations.jsonl, and prompts the user to test.Run the meta-optimize skill after at least 5 logged workflow runs by using the command "/meta-optimize" to generate recommended SKILL.md and prompt optimizations.
High-intent search queries and answers about installing and using this skill.
To optimize skill prompts, you analyze usage logs to identify failure patterns and generate evidence-backed patches for SKILL.md files. This process requires reading event logs and proposing minimal diffs with frequency and convergence analysis.
Log-driven prompt tuning is the process of analyzing recorded workflow events to identify parameter overrides and failures, then proposing minimal patches to workflow defaults and review prompts to improve system performance.
You need at least 5 logged workflow runs before analyzing usage records for skill optimization. The system requires sufficient data in the events log to perform meaningful frequency, failure, and human-intervention analysis.
No, skill optimization never auto-applies patches. It backs up changes, runs a cross-model reviewer loop, records an optimization trail, and requires explicit user approval before applying any proposed diffs to SKILL.md files.
The best way to identify failing parameters is performing frequency and convergence analysis on event logs, specifically looking for correlations between workflow defaults, overrides, and human interventions recorded during execution.
Workflow defaults stop matching actual usage because review prompts and parameters degrade when user behavior evolves. Analyzing event logs for override patterns and generating minimal evidence-backed patches realigns the system with current usage.
duypham2801Analyze JSONL event logs and propose reviewer-gated SKILL.md patches.
goupup-aiAnalyze ARIS usage logs to propose evidence-backed SKILL.md patch improvements.
Mang30Analyze ARIS usage logs to propose evidence-backed SKILL.md patch diffs.
CharlieGreenmanConfigure GPD workflow settings interactively and merge them into .gpd/config.json.