Developer Tools●●●Banger
Sift, a small CLI that groups noisy test failures into root causes
Compresses 198k tokens to 129 by grouping test failures before the agent sees them.
Big BrainSolve My Problem
bimamoglu
202mo ago
Back to browse
GitHub Repository
Turn noisy command output into a short, actionable first pass for coding agents.
62 starsTypeScript
Open-source CLI that turns 128 test failures into 2 root causes
by bimamoglu·Mar 19, 2026·2 points·1 comment
AI Analysis
●●SolidNiche GemShip It
Heuristic-first parsing cuts 198K tokens to 129 before the LLM ever sees output.
Strengths
- •Local heuristics group failures before LLM escalation — smart cost optimization for token-heavy workflows.
- •Returns structured diagnosis with anchor file, likely fix, and decision signal for agents.
- •Works with pytest, typecheck, lint, build, and audit workflows out of the box.
Weaknesses
- •Narrow scope — only useful for developers running AI agents against noisy test output.
- •Early stage with 12 stars — needs more real-world validation beyond the author's benchmarks.
Category
Target Audience
Developers using AI coding agents for debugging
Similar To
pytest-rich · pytest-sugar
Post Description
I kept running into the same issue with coding agents.
A test run fails, you get a huge wall of output, and most of the effort goes into figuring out what actually went wrong.
In many cases, the failures are not independent. It’s the same issue repeated across many tests.
In one case: 128 failures → 2 root causes
I built a small CLI that groups repeated failures into shared root causes before passing the result to the model.
It’s mainly built for coding agents, but works on raw CLI output as well.
On my backend tests, this reduced debugging time and token usage quite a bit.
Similar Projects
Developer Tools●●●Banger
Sift, a local-first CLI for failures, root causes and next steps
198k tokens down to 129 — local heuristics beat LLM summarization.
Solve My ProblemBig BrainShip It
bimamoglu
312mo ago
Infrastructure●●Solid
Kroot – dependency-graph root cause analysis for Kubernetes
Kubernetes root cause via dependency graphs, but kubectl debug and observability tools already solve this.
Solve My ProblemNiche Gem
An0n_Jon
113mo ago
Developer Tools●●Solid
PipelineIQ – AI diagnoses CI/CD failures and sends fixes to Slack
Claude diagnoses pipeline failures in seconds, but log analysis tools already exist.
Solve My Problem
Raja_Dev
103mo ago
AI/ML●●●Banger
Kelet – Root Cause Analysis agent for your LLM apps
Auto-clusters failure patterns across sessions and suggests prompt patches.
Solve My ProblemBig Brain
almogbaku
47242mo ago
Developer Tools●●Solid
Nightwatch, The open-source, read-only AI SRE
Read-only AI agent architecture prevents production accidents during incident response.
Big BrainShip It
egorferber
3397d ago