Skip to content

[observability escalation] Design Decision Gate πŸ—οΈ β€” repeated resource-heavy runs with increasing turn countsΒ #27293

Closed as not planned

Description

@github-actions

Design Decision Gate πŸ—οΈ has crossed the escalation threshold with 2 consecutive runs both carrying a high-severity resource_heavy_for_domain assessment, cohort-matched baselines confirming turns_increase, and the latest run also showing medium poor_agentic_control.

Evidence

Run Turns Classification Assessments
Β§24648171195 4 baseline β€”
Β§24649877750 8 changed / turns_increase resource_heavy_for_domain HIGH
Β§24650414222 10 changed / turns_increase resource_heavy_for_domain HIGH, poor_agentic_control MEDIUM

The turn count has grown 4 β†’ 8 β†’ 10 across 3 runs today. The latest run used 9 distinct tool types in an exploratory, broad-tool posture. Agentic fraction is ~50%, meaning roughly half the turns are pure data-gathering that could be moved to deterministic pre-steps.

Why This Matters

  • Engine: claude (frontier model cost)
  • Execution style: exploratory, tool breadth: broad β€” weakly scoped
  • The system flagged it as partially_reducible and model_downgrade_available
  • Pattern is worsening across runs, not stabilizing

Recommended Fix

  1. Tighten instructions β€” Reduce scope so the agent does not explore broadly. Narrow the task to a specific question rather than open-ended analysis.
  2. Move data-gathering to pre-steps β€” The ~50% data-gathering fraction can become deterministic frontmatter steps writing to /tmp/gh-aw/agent/, reducing inference turns.
  3. Reduce tool breadth β€” Limit available tools to what the task strictly needs. 9 tool types for an issue-response workflow is excessive.
  4. Try a smaller model β€” claude-haiku-4-5 is recommended by the assessment. The task (issue response / decision gating) is well-suited to a smaller model.

Suggested Route

workflow:Design Decision Gate πŸ—οΈ β€” workflow owners should review the prompt and tool configuration.

See the full observability report discussion for context on related workflows and the broader resource-heaviness pattern across the repository.

Generated by Agentic Observability Kit Β· ● 887.4K Β· β—·

Activity

  1. github-actions commented on Apr 21, 2026

    @github-actions
    ContributorAuthor

    This issue is being closed as outdated. A newer issue has been created: #27637

    View newer issue


    This action was performed automatically by the Agentic Observability Kit workflow.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions