agentmods.dev

mushi-debugger

mushi-debugger is an agent for coding agents from kensaurus/mushi-mushi. It costs 79 tokens per session (595 once invoked), scanned A, original, MIT.

An AI coding-agent role that turns Mushi user bug reports into code changes. It retrieves a report’s diagnosis, impact, reproduction steps, and prepared fix instructions before editing the repository.

In plain words
What is it for?
Use it to choose a report to fix, locate the affected code, make a focused change, run tests, submit the result, or ask Mushi to dispatch its own fix attempt.
Why use it?
It gives the agent structured context before code changes, reducing the chance of fixing only the visible symptom. It also applies project-specific lessons, runs relevant tests, and links the result back to the report.

Agent

Part of the mushi-debugger plugin — 1 agent shipped together

not rated 3 6d ago A scan 79 tokens original MIT
View the source file kensaurus/mushi-mushi
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/kensaurus/mushi-mushi/mushi-debugger
Clone the repo
git clone --depth 1 https://github.com/kensaurus/mushi-mushi

Or install mushi-debugger, the plugin that ships this one along with the rest of its 1 agent.

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 mushi-debugger

README.md
[![agentmods](https://agentmods.dev/badge/agents/kensaurus/mushi-mushi/mushi-debugger.svg)](https://agentmods.dev/agents/kensaurus/mushi-mushi/mushi-debugger)
Your own site
<a href="https://agentmods.dev/agents/kensaurus/mushi-mushi/mushi-debugger"><img src="https://agentmods.dev/badge/agents/kensaurus/mushi-mushi/mushi-debugger.svg" alt="Measured on agentmods" height="20"></a>
Per session 79 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 595 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 $0.00079 $0.00595
Opus 5 $0.00039 $0.00298
Sonnet 5 $0.00016 $0.00119
Haiku 4.5 $0.00008 $0.00060

Measured 3d ago against content hash 6f49c952d5fe, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

mushi-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 3d 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.

plugins/mushi-debugger/agents/mushi-debugger.md · 51 lines

What it actually says

You are the Mushi debugger. Your job: turn a user bug report into a shipped fix using the mushi MCP tools plus this repo's code.

Workflow

  1. Find the report. If the user named one, call get_report_detail. Otherwise call get_recent_reports (filter by status new/triaged) or triage_issue for a "what should I fix next" ranking. search_reports handles fuzzy descriptions ("the checkout crash").

  2. Get the fix brief before touching code. Call get_fix_context for the chosen report — it returns root cause analysis, blast radius, repro steps, and a server-composed fix prompt. Also call query_lessons to apply any project lessons that match the affected area. Never start editing from the raw report text alone.

  3. Locate and fix. Map the brief's file/function references into this repo (Grep/Glob), make the smallest change that resolves the root cause, and run the repo's tests for the touched area.

  4. Close the loop. Call submit_fix_result with the branch/PR and files changed so the report links to the fix and the judge can score it. If the user wants Mushi's own agent to do the work instead, call dispatch_fix and report back the dispatch status.

  5. Reply to the reporter when asked. reply_to_reporter sends a plain-language update to the person who filed the bug.

Rules

  • Diagnosis quota is metered: get_fix_context on an already-diagnosed report is free; avoid forcing re-classification unless evidence changed.
  • Prefer fixing in this repo over dispatch_fix when the change is small — dispatch is for when the user explicitly wants it hands-off.
  • Never mark a report resolved without either a merged fix or the user's explicit say-so; use transition_status honestly.
  • If the MCP server is not connected, tell the user to run /mcp and sign in to mushi (browser OAuth — no API key needed), or npx mushi-mushi setup --ide claude from the terminal.
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. 3d ago First seen · 51 lines · 79 tokens per session scan A 6f49c952d5fe

Subscribe to this mod's changes

mushi-debugger is an agent published in the GitHub repository kensaurus/mushi-mushi (3 stars, last pushed 6d ago), licensed MIT. It adds 79 tokens to every session and 595 once invoked, about $0.0004 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.

Debuggingangularauto-fixbug-reportingclaude-codefeedback-widgetllm-ops
Related

Other agents, from other repositories

ci-watcher

Polls Nx Cloud CI pipeline and self-healing status. Returns structured state when actionable. Spawned by /nx-cloud-ci-monitor command to monitor CI Attempt status.

nrwl/nx · 37 tokens

nx-workflow

Purpose: Understanding the NX monorepo structure, development patterns, and best practices. For commands, see docs/internal/commands.md.

SAP/fundamental-ngx · 0 tokens

testing-guidelines

Agent "testing-guidelines" from SAP/fundamental-ngx, covering testing guidelines, test commands, run all tests in a library, run specific test file and examples.

SAP/fundamental-ngx · 0 tokens

instruments-misc

Instruments outside the obc-watch core — readout rows, propulsion glyphs, and helper-borrowing SVG instruments.

Ocean-Industries-Concept-Lab/openbridge-webcomponents · 29 tokens

docgen

The OpenAI-backed JSDoc generation CLI and its review-copy workflow.

Ocean-Industries-Concept-Lab/openbridge-webcomponents · 17 tokens

a11y-write-playwright-tests

Write Playwright accessibility tests for a component or page: axe scan file and keyboard-only variant. Triggers when the user asks to 'add a11y tests', 'write axe test', 'add keyboard test', or 'create accessibility test for X'.

AmadeusITGroup/otter · 59 tokens