neuron-nki-debugger-agent

neuron-nki-debugger-agent is an agent for Claude Code from aws-neuron/neuron-agentic-development. It costs 239 tokens per session (2,757 once invoked), scanned A, original, Apache-2.0.

An automated helper for debugging NKI kernel compilation errors. NKI is a language for writing computation kernels that run on AWS Neuron chips, such as Trainium and Inferentia.

In plain words
What is it for?
Use it when an NKI kernel fails to compile and needs code corrections that follow the current Neuron SDK rules.
Why use it?
It helps interpret compiler errors, find relevant documentation or examples, apply straightforward fixes, and check whether the corrected kernel compiles.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: model in frontmatter; names the TodoWrite tool.

Good fit Use it when an NKI kernel fails to compile and needs code corrections that follow the current Neuron SDK rules.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-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.

Clone the repo
git clone --depth 1 https://github.com/aws-neuron/neuron-agentic-development

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for neuron-nki-debugger-agent

README.md
[![agentmods](https://agentmods.dev/badge/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent/github.svg)](https://agentmods.dev/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent)
Your own site
<a href="https://agentmods.dev/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent"><img src="https://agentmods.dev/badge/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for neuron-nki-debugger-agent

Your own site · 80×15
<a href="https://agentmods.dev/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent"><img src="https://agentmods.dev/badge/agents/aws-neuron/neuron-agentic-development/neuron-nki-debugger-agent.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 239 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 2,757 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00239 $0.02757
Opus 5.5 $0.00096 $0.01103
Sonnet 5.5 $0.00048 $0.00551
Haiku 4.5 $0.00024 $0.00276

Measured 1mo ago against content hash 4ab647106f21, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

neuron-nki-debugger-agent 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 1mo 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.

agents/neuron-nki-debugger-agent.md · 343 lines

How it starts

The opening of the file, as written. The whole thing — 343 lines — stays where its author put it; the contents beside it link to each section on GitHub.

NKI Debugger Agent

You are an expert NKI kernel debugger. Your role is to autonomously debug and fix NKI kernel compilation errors through pragmatic error analysis, documentation lookup, and incremental fixes following the principle of simplicity over performance.

NKI Language Constraints (MANDATORY)

CRITICAL: All NKI code you generate MUST follow the language constraints defined in /neuron-nki-writing reference nki-language-constraint.md. Code that violates these constraints will NOT compile on current Neuron SDK.

Read /neuron-nki-writing reference nki-language-constraint.md for the full constraint table and reference kernel. If you cannot load the skill, follow the reference kernel in the description examples above.

Debugging Philosophy

Follow these core principles in order:

  1. Obvious fixes first - If the fix is clear from the compiler error message, apply it immediately
  2. Additional context - Check for additional compiler error messages that can help
  3. Learn from examples - Look for code examples doing similar tasks for alternative implementations
  4. Simplicity over performance - Sacrifice performance and use simpler patterns when needed

CRITICAL: When making performance trade-offs for simplicity, ALWAYS document the trade-off explicitly in your report so the user understands what was sacrificed.

Debugging Workflow

Execute these phases in order. Track iterations to prevent infinite loops (max 10 iterations).

Phase 1: Analyze Error Message

  1. Run compilation to capture the full error output:
source $NKI_VENV_PATH/bin/activate
python test_{kernel_name}.py
  1. Parse error information:

    • Error code (NCC_EVRF*, NCC_EOOM*, NCC_EARG*, NCC_EHCA*, etc.)
    • Line number and operation name
    • Error description and context
    • Any suggestions in the error message
  2. Create error analysis in your report:

## Error Analysis

**Error Code:** {error_code}
**Category:** {Verification | Memory | Type/Operation | etc.}
**Location:** Line {line_num}, {function_name}()
**Issue:** {description}
**Compiler Suggestion:** {if any}

Read the full file on GitHub · 343 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. 1mo ago First seen · 343 lines · 239 tokens per session scan A 4ab647106f21

Subscribe to this mod's changes

neuron-nki-debugger-agent is an agent published in the GitHub repository aws-neuron/neuron-agentic-development (63 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 239 tokens to every session and 2,757 once invoked, about $0.0010 per session on Opus 5.5. 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 2026-08-30.

Related

Other agents, from other repositories

network-troubleshooter

Diagnoses network connectivity, routing, DNS, interface, and policy symptoms with a read-only OSI-layer workflow and evidence-backed root cause summary.

affaan-m/ECC · 34 tokens

waveform-analyzer

VCD/FST deep analysis specialist. Root causes protocol violations and traces multi-clock signal relationships. Never modifies RTL or test files.

babyworm/rtl-agent-team · 31 tokens

tdmcp-implementation-runtime-analyst

Studies the lived runtime path of a completed tdmcp implementation: TouchDesigner bridge behavior, hardware setup, diagnostics, calibration, audio/video/projector issues, latency, and user-facing operator ergonomics.

Pantani/tdmcp · 50 tokens

bringup-debugger

Use this agent when bare-metal code, firmware, or an FPGA design will not come up on real hardware: it boots but hangs, faults early, gives no output, or configuration silently fails. It drives the bring-up checkpoint ladder and the known per-layer gotchas to localize which rung is broken. Dispatch it for "boots in…

Midstall/claude-for-hardware · 80 tokens

jetson-perf-investigator

Diagnose Jetson performance and memory issues end to end. Captures a baseline, isolates the dominant consumer, and recommends concrete next actions by orchestrating the Jetson Platform Skills catalog. Use when the user reports "Jetson is slow / hot / OOM" without a clear cause.

NVIDIA-AI-IOT/jetson-device-skills · 65 tokens

bubbaloop-operator

Use when the user asks to inspect, control, or troubleshoot a running Bubbaloop deployment — node lifecycle (start/stop/restart/build), health checks, log reads, mission management, or scheduling tasks. Wraps the bubbaloop MCP server and enforces the discover-before-act discipline.

kornia/bubbaloop · 67 tokens