debug

A debugging command that investigates a bug, failing test, or unexpected behavior by finding the underlying cause, making a small fix, and adding a regression test. A regression test checks that the same problem does not return.

In plain words
What is it for?
Use it to diagnose application bugs, broken tests, and unexpected results, optionally using a ticket reference for added context.
Why use it?
It provides a repeatable investigation path and helps prevent a fixed bug from being reintroduced. Slow or latency-related problems are sent to a separate performance workflow.

Command

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 commands/lugassawan/swe-workbench/debug
Clone the repo
git clone --depth 1 https://github.com/lugassawan/swe-workbench
Per session 46 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,975 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 $0.00046 $0.01975
Opus 5 $0.00023 $0.00988
Sonnet 5 $0.00009 $0.00395
Haiku 4.5 $0.00005 $0.00198

Measured yesterday against content hash 96f7c3b86310, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

debug 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 yesterday.

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.

commands/debug.md · 107 lines

How it starts

The opening of the file, as written. The whole thing — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Symptom: $ARGUMENTS

If $ARGUMENTS contains a ticket reference, invoke swe-workbench:ticket-context first and prepend its structured summary to the delegation context below. Skip if $ARGUMENTS is free-text with no recognizable ref. (Trigger patterns are defined in that skill's "When to invoke" section.)

Interrogation mode. Before producing anything, resolve the mode:

  • Explicit signal in the invocation is honored without asking. grill-me = --grill, "grill me", or "grill-me mode". standard = --standard, "standard", or "quick". Strip the signal from $ARGUMENTS and record the resolved mode.
  • No explicit signal: ask via AskUserQuestion — one question, header "Mode", options Standard (recommended, listed first) and Grill me. Standard description: "Lightweight clarify — a restatement and at most one question, then proceed." Grill-me description: "Relentlessly walk the decision tree one question at a time, each with a recommended answer, self-answering from the codebase where possible." Use the user's choice.

Standard mode: proceed with the command's existing lightweight clarify (a restatement and at most one clarifying question) — do not ask the mode question again.

Grill-me mode: activate swe-workbench:workflow-grill and run its interrogation loop to completion (exit on shared understanding or when the user says "proceed"). Then thread the emitted ## Resolved decisions block into the command's normal artifact/delegation step below — the same way a ticket-context summary is prepended — and continue as in standard mode.

Performance routing. After resolving the mode, check whether the symptom is performance-shaped before proceeding. A symptom is performance-shaped when it describes speed, latency, throughput, or resource exhaustion without an assertion failure or incorrect output — key vocabulary (case-insensitive, stem-matched): slow, too slow, latency, throughput, cpu, memory leak, oom, profil*, benchmark, bottleneck, perf, performance, high memory, resource exhaustion, takes too long.

If the symptom is performance-shaped:

  1. If grill-me mode was active and the ## Resolved decisions block already classifies the symptom as performance or correctness, use that classification and skip the AskUserQuestion.

  2. Otherwise call AskUserQuestion:

    {
      "questions": [
        {
          "question": "Is this a performance investigation or a correctness bug?",
          "header": "Symptom type",
          "multiSelect": false,
          "options": [
            {
              "label": "Performance investigation",
              "description": "Activate `swe-workbench:workflow-performance-investigation`: profile-first runbook (baseline → profile → ranked hotspots → one change → before/after measurement → regression guard)."
            },
            {
              "label": "Correctness bug",
              "description": "Proceed with the debugger: root-cause first, minimal fix, regression test."
            }
          ]
        }
      ]
    }
    
  3. On Performance investigation: activate swe-workbench:workflow-performance-investigation and stop — do not continue to browser diagnostics or debugger delegation.

  4. On Correctness bug: continue with the browser diagnostic step below.

If the symptom is not performance-shaped: skip this step entirely.

Browser diagnostic step (web-UI symptoms only). Before delegating, check whether the symptom indicates a browser / web-UI context (e.g., mentions a rendered page, console error, XHR/fetch failure, visual regression, or DOM state). If it does:

  1. Gate — check whether a Chrome backend is available in this session. A Chrome backend is one of:

    • chrome-devtools-mcp (read_console_messages, read_network_requests tools reachable), OR
    • Claude-in-Chrome (mcp__claude-in-chrome__read_console_messages tool reachable).

    If the symptom is web-UI but no Chrome backend is connected, stop immediately and return:

Read the full file on GitHub · 107 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. yesterday First seen · 107 lines · 46 tokens per session scan A 96f7c3b86310

Subscribe to this mod's changes

debug is a command published in the GitHub repository lugassawan/swe-workbench (2 stars, last pushed yesterday), licensed MIT. It adds 46 tokens to every session and 1,975 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.