systematic-debugging

A four-step method for finding the real cause of a software bug: observe the problem, form explanations, test them, and then fix the confirmed cause. It also requires a small reproducible example and evidence that the fix works.

In plain words
What is it for?
Use it for incidents, flaky tests, bugs spanning multiple modules, or any problem where the cause is unclear.
Why use it?
It reduces random patching and helps avoid fixes that appear to work without explaining or preventing the original failure.

Skill for Claude CodeCodex

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 skills/kbwen/agentic-os/systematic-debugging
Any agent
npx skills add KbWen/agentic-os --skill systematic-debugging
Clone the repo
git clone --depth 1 https://github.com/KbWen/agentic-os

Made for: Claude Code, Codex.

Per session 28 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 431 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 $0.00028 $0.00431
Opus 5 $0.00014 $0.00216
Sonnet 5 $0.00006 $0.00086
Haiku 4.5 $0.00003 $0.00043

Measured 2d ago against content hash 62c160464493, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

systematic-debugging 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/skills/systematic-debugging/SKILL.md · 57 lines

What it actually says

Systematic Debugging

Overview

The core of systematic debugging is: understand first, then fix. When encountering a bug, clarify symptoms, reproduction conditions, and blast radius. Draw hypotheses, verify them with experiments to isolate the root cause, and only then submit a minimal, verifiable fix.

Ironclad Rules

  1. No random patching: Do not submit fixes without root cause evidence.
  2. Change one variable at a time: Prevent unexplainable results from touching multiple areas at once.
  3. Fixes MUST include evidence: Include reproduction steps, verification, and regression results.

When to Use

  • Hotfix incident response.
  • Flaky tests.
  • Cross-module anomalies that aren't intuitively obvious.
  • Any "fixed but I don't know why" risk scenarios.

Four-Phase Process

Phase 1: Observe

  • Precisely record error messages, timestamps, and input conditions.
  • Create a Minimal Reproducible Example (MRE).
  • Mark the blast radius (affected modules/users).

Phase 2: Hypothesize

  • Propose 1–3 testable root cause hypotheses.
  • Design "falsifiable" checks for each hypothesis.
  • Prioritize high-probability, low-cost verifiable items.

Phase 3: Verify

  • Run experiments and retain output logs.
  • Adjust only one variable to confirm causality.
  • Remove falsified hypotheses to converge on the most likely root cause.

Phase 4: Fix

  • Implement a Minimal Fix.
  • Add a Regression Test.
  • Verify: The original error disappears AND existing behavior is not broken.

Common Mistakes

  • Modifying code before reproducing the issue.
  • Modifying too many files at once, failing to locate the effective fix point.
  • Treating "accidental passes" as root cause resolved.
  • Lacking regression tests, causing similar issues to happen again.
Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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 · 57 lines · 28 tokens per session scan A 62c160464493

Subscribe to this mod's changes

systematic-debugging is a skill published in the GitHub repository KbWen/agentic-os (149 stars, last pushed 8d ago), licensed MIT. It adds 28 tokens to every session and 431 once invoked, about $0.0001 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 2026-08-30.

Related

Other skills, from other repositories

orloj-generator

Interactive scaffold generator for Orloj multi-agent systems. Use this skill whenever someone wants to create, set up, scaffold, bootstrap, or generate an Orloj agent system, pipeline, swarm, or hierarchy. Also trigger when users mention "orlojctl init", ask how to get started with Orloj, want to build a multi-agent…

OrlojHQ/orloj · 168 tokens

muster-frappe-metadata

Design, generate, review, apply, and roll back Frappe or ERPNext metadata changes through Muster. Use for DocTypes, child tables, custom fields, property setters, workflows, roles, permissions, reports, print formats, web pages, workspaces, fixtures, patches, or installed-app compatibility.

musterhq/muster · 69 tokens

muster-frappe-release

Build, seed, deploy, break, measure, and prove a Muster Frappe release. Use for bench or site setup, migrations, realistic ERPNext data, role matrices, browser/mobile QA, performance and recovery tests, release evidence, Frappeverse demos, or normal-speed presentation video.

musterhq/muster · 66 tokens

muster-frappe-security

Threat-model and negatively verify Muster integrations with Frappe OAuth, multi-tenancy, RBAC, field permissions, agent delegation, channels, secrets, approvals, and audit logs. Use for identity linking, site registration, Telegram onboarding, policy changes, permission tests, or security release gates.

musterhq/muster · 64 tokens

muster-frappe-workflows

Design and implement durable multi-step Muster workflows for Frappe with agent, subagent, and nested task graphs. Use for workflow definitions, planning, delegation, tool policies, approvals, retries, schedules, triggers, artifacts, dynamic page creation, or Claude-style universal JS and JSON automation.

musterhq/muster · 65 tokens

muster-frappe-orchestrator

Coordinate end-to-end Muster automation work across Frappe or ERPNext architecture, metadata, UI, workflows, security, testing, deployment, and evidence. Use for cross-cutting features, production plans, multi-agent delivery, or work that spans Muster core and a Frappe app.

musterhq/muster · 66 tokens