Output filtering causes AI agents (Claude Code) to burn extra tokens on retry loops

[BadassBison]

#791 (fix)
#871 (doc updates)

[pszymkowiak]

[w] wshm · Automated triage by AI

🔍 Automated Triage

🐛 Category	bug
🟠 Priority	high
🎯 Confidence	82%

Summary

Output filtering strips information that AI agents need, causing them to retry commands and waste tokens; linked PR #791 likely contains the fix.

---

Triaged automatically by wshm · This is an automated analysis, not a human review.

[krabat-l]

Interesting connection between output fidelity and agent behavior. Even with full unfiltered output though, agents often still loop because they skip investigation and jump straight to editing.

A complementary approach: hooks that block Edit/Write calls during investigation phases, forcing the agent to reproduce and isolate before fixing. This addresses the workflow discipline side while RTK addresses the information fidelity side.

Open-source implementation for Claude Code: claude-debug