[perf-improvement] AIC Budget Crisis Day 5 — 6-agent cluster expanding, root fix urgently needed

[github-actions]

Problem

The max-ai-credits: 1000 guardrail has been blocking analysis-heavy workflows for 5 consecutive days (Jun 9–13, 2026). Today the cluster expanded from 5 to 6 affected agents, with no root fix applied.

Affected Workflows — June 13, 2026

Workflow	Issue	Failure Mode
Code Simplifier	#39013	HTTP 429 rate-limited (provider quota exhausted from yesterday’s 4,219 AIC runaway)
Test Quality Sentinel	#39059	AIC budget exceeded (1000 limit)
Matt Pocock Skills Reviewer	#39050	AIC budget exceeded (daily limit)
Daily CLI Tools Exploratory Tester	#39031	AIC rate limit
jsweep (JavaScript Unbloater)	#39020	Tool denial limit (AIC exhaustion pattern)
Copilot CLI Deep Research	#39022	Tool denial limit

History

• Jun 9–11: 4 agents blocked — original cluster tracker [ai-credits-cluster] AI Credits Exhaustion Cluster — 2026-06-11: 4 workflows failing #38520 (closed)
• Jun 11: Specific tracker filed for Failure Investigator ([perf-improvement] Raise max-ai-credits for Failure Investigator (6h) — meta-monitor blind spot #38624, open)
• Jun 12: 5 agents blocked + Code Simplifier runaway (4,219 AIC consumed)
• Jun 13: 6 agents blocked — this issue

Root Cause

Analysis-heavy workflows have max-ai-credits: 1000 or hit the 24h rolling AIC budget. As the repository grows and workflows handle increasingly complex analysis tasks, 1000 AIC is insufficient for a full analysis run.

The Code Simplifier’s Jun 12 runaway (4,219 AIC from a bash permission loop) further exhausted provider quota, causing HTTP 429 rate limiting today.

Recommended Fixes

1. Raise max-ai-credits from 1000 to 2000 for the following workflows:

   • Code Simplifier
   • Test Quality Sentinel
   • Matt Pocock Skills Reviewer
   • Daily CLI Tools Exploratory Tester
   • Any other analysis-heavy workflows currently hitting the 1000 limit

2. Apply a runaway guard to Code Simplifier: max-turns: 30, bash tool allowlist, max-ai-credits: 1500 (specific fix tracked in [perf-improvement] Raise max-ai-credits for Failure Investigator (6h) — meta-monitor blind spot #38624)

3. Implement soft pre-cap warning (already recommended in [deep-report] Add a soft pre-cap AI-credits guard to heavy aggregator workflows #38645): warn at 75% AIC usage before hard failure

4. Fix jsweep and Copilot CLI Deep Research tool denial: Review and expand allowed tool list or add retry with scoped tools

Impact

• 6 agents generating 0 useful output per run while consuming quota
• Code Simplifier’s rate-limit cascades affect other workflows sharing the provider quota
• Analysis coverage gap: test quality, code simplification, CLI tooling all dark

Urgency

This pattern has persisted for 5 days without a root fix. The cluster is expanding (+1 agent per day on average). Without intervention, additional agents will be added to the cluster daily.

References:

• Cluster history: [ai-credits-cluster] AI Credits Exhaustion Cluster — 2026-06-11: 4 workflows failing #38520 (Jun 11, closed), [ai-credits-cluster] AI Credits Exhaustion Cluster Expanding: 8 Analysis-Heavy Workflows Failing #38119 (earlier, closed)
• Related fixes: [perf-improvement] Raise max-ai-credits for Failure Investigator (6h) — meta-monitor blind spot #38624, [deep-report] Add a soft pre-cap AI-credits guard to heavy aggregator workflows #38645
• Today’s per-agent issues: [aw] Code Simplifier failed #39013, [aw] Test Quality Sentinel exceeded daily AI credits budget #39059, [aw] Matt Pocock Skills Reviewer exceeded daily AI credits budget #39050, [aw] Daily CLI Tools Exploratory Tester hit AI credits rate limit #39031, [aw] jsweep - JavaScript Unbloater exceeded tool denial limit #39020, [aw] Copilot CLI Deep Research Agent exceeded tool denial limit #39022

Generated by ⚡ Agent Performance Analyzer - Meta-Orchestrator · 608.5 AIC · ⌖ 21.6 AIC · ⊞ 22.2K · ◷

• expires on Jun 15, 2026, 5:23 AM UTC-08:00

[copilot-swe-agent]

The agent encountered an error and was unable to start working on this issue: Please try again later, or contact support if the problem persists. (Request id: 1A40:3FC6F0:22A4E2D:80D7EAF:6A2D5EE2)

[github-actions]

🍪 Issue Monster selected this for Copilot

I've identified this issue as a good candidate for automated resolution and requested assignment to the Copilot coding agent.

If assignment succeeds, the Copilot coding agent will analyze the issue and create a pull request with the fix.

Om nom nom! 🍪

🍪 Om nom nom by Issue Monster · 17.8 AIC · ⌖ 4.37 AIC · ⊞ 22.3K · ◷