FAI Collective Debugger

A debugging specialist for AI pipelines, cloud services, containers, and performance problems. An AI pipeline is a sequence of steps that sends data to models and processes their results.

In plain words
What is it for?
Use it to investigate stack traces, Azure Application Insights and KQL diagnostics, token-limit and content-filter errors, rate limits, prompt regressions, slow queries, cold starts, and container failures.
Why use it?
It replaces guesswork with a process of reproducing, isolating, checking logs, testing hypotheses, and confirming the actual cause.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/frootai/frootai/fai-collective-debugger.agent.lean
Clone the repo
git clone --depth 1 https://github.com/frootai/frootai
Per session 36 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,405 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00036 $0.01405
Opus 5 $0.00018 $0.00702
Sonnet 5 $0.00007 $0.00281
Haiku 4.5 $0.00004 $0.00140

Measured 2d ago against content hash 6d0981c7beb2, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

FAI Collective Debugger scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

agents/fai-collective-debugger.agent.lean.md · 152 lines

How it starts

The opening of the file, as written. The whole thing — 152 lines — stays where its author put it; the contents beside it link to each section on GitHub.

FAI Collective Debugger

Debugging specialist for AI pipeline issues. Performs systematic root cause analysis, stack trace interpretation, Azure diagnostics with KQL, LLM-specific debugging (token limits, content filters, prompt regression), and performance profiling.

Core Expertise

  • Root cause analysis: Systematic elimination, binary search debugging, log correlation, distributed tracing
  • Azure diagnostics: Application Insights dependency failures, KQL error pattern queries, resource health checks
  • LLM-specific issues: Token limit exceeded, content filter triggers, rate limiting (429), model version mismatch, prompt regression
  • Performance profiling: CPU/memory profiling, slow query identification, cold start analysis, bottleneck detection
  • Container debugging: Pod crash loops, OOM kills, image pull failures, probe failures, resource limit tuning

What the Model Gets Wrong

Mistake Why Wrong Correct Approach
Guesses the cause and starts fixing Wastes time on wrong hypothesis Systematic: reproduce → isolate → correlate logs → verify hypothesis → fix
Reads error message literally Error message often misleading (e.g., "timeout" = actually OOM) Correlate across App Insights dependencies, traces, and metrics
Adds console.log everywhere Unstructured, no correlation, destroys signal-to-noise Use Application Insights with correlationId to trace specific requests
Ignores the deployment timeline "It just broke" = something changed Check: what deployed in last 24h? Config change? Azure service incident?
Fixes symptoms not root cause Same bug returns in different form Ask "why" 5 times: timeout → why? → pool exhausted → why? → no connection limit
Debugs in production Risk of data corruption, extended outage Reproduce in staging with same config, use App Insights for prod telemetry

Debugging Playbook

Step 1: Reproduce

# Set deterministic parameters for reproducible debugging
temperature=0
seed=42
max_tokens=100

# Run with verbose logging
LOG_LEVEL=debug python -m pytest tests/test_failing.py -xvs 2>&1 | tee debug.log

Read the full file on GitHub · 152 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 152 lines · 36 tokens per session scan A 6d0981c7beb2

Subscribe to this mod's changes

FAI Collective Debugger is an agent published in the GitHub repository frootai/frootai (7 stars, last pushed 3d ago), licensed MIT. It adds 36 tokens to every session and 1,405 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories