Coordinate parallel agent-eval and orch-eval workflows to grade agent sessions.
npx skills add https://github.com/EdgeApp/edge-dev-agents --skill eval-run-edgeapp
Please help me install this Agent Skill. Skill: eval-run Source: https://github.com/EdgeApp/edge-dev-agents/tree/main/.cursor/skills/eval-run Command: npx skills add https://github.com/EdgeApp/edge-dev-agents --skill eval-run-edgeapp
This skill solves the challenge of verifying and scoring the performance of autonomous agent runs, ensuring that complex development workflows meet quality standards without manual oversight.
Invoke the eval-run skill to evaluate all agent runs from yesterday using the default report-based mode.
High-intent search queries and answers about installing and using this skill.
Automate quality assurance by orchestrating parallel agent-eval and orch-eval workflows to grade Asana-backed development tasks against defined rubric dimensions and infrastructure gates.
Adversarial verification automatically re-verifies negative findings during agent performance grading to ensure accuracy and prevent false positives in the evaluation reports.
Evaluate autonomous agent sessions by invoking the dual-mode evaluation feature, which supports both lightweight report-based grading and deep-dive transcript analysis.
Yes, agent workflow evaluation requires the jq command-line dependency to be installed in your environment to parse and process the evaluation data.
Yes, you can generate actionable remediation checklists containing approved drafts to fix workflow gaps or promote successful patterns directly from the evaluation results.
The best way to synthesize cohort-level patterns is by coordinating parallel evaluation workflows that systematically grade Asana-backed development tasks against defined rubric dimensions.
EdgeAppResolve agent sessions into structured JSON evidence manifests with transcripts and logs.
takemo101Automate post-PR workflows by monitoring CI checks, fixing failures, and merging.
quanganh208Create or update Claude skills with eval-driven iteration and bundled resources.
lehuuhau1231Analyze source code against structured checklists for quality, security, and performance.
