Asynkron.JsEngine: Agent for Claude Code

.claude/agents/deep-thinker.md

deep-thinker is an agent for Claude Code from asynkron/Asynkron.JsEngine. It costs 387 tokens per session (1,525 once invoked), scanned A, original, MIT.

A debugging agent for complex bugs and failing tests in .NET JavaScript engine code. It gathers detailed test output, logs, stack traces, and profiling information before forming an explanation.

In plain words
What is it for?
Use it to investigate failing tests, reproduce engine errors, inspect diagnostic logs, and produce actionable findings.
Why use it?
It replaces broad guessing with evidence from targeted test runs and internal diagnostics. This helps separate parser, runtime, asynchronous, iterator, and other possible causes.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter.

This is asynkron/Asynkron.JsEngine's own configuration. It tells Claude Code how to work on Asynkron.JsEngine itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything Asynkron.JsEngine configures →

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is ./tools/profile <script-name> --cpu.

Reuse

Borrowing it

Nothing to install: this file belongs to asynkron/Asynkron.JsEngine. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/asynkron/Asynkron.JsEngine/main/.claude/agents/deep-thinker.md
Clone the repo
git clone --depth 1 https://github.com/asynkron/Asynkron.JsEngine

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 deep-thinker

README.md
[![agentmods](https://agentmods.dev/badge/agents/asynkron/asynkron.jsengine/deep-thinker/github.svg)](https://agentmods.dev/agents/asynkron/asynkron.jsengine/deep-thinker)
Your own site
<a href="https://agentmods.dev/agents/asynkron/asynkron.jsengine/deep-thinker"><img src="https://agentmods.dev/badge/agents/asynkron/asynkron.jsengine/deep-thinker/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 deep-thinker

Your own site · 80×15
<a href="https://agentmods.dev/agents/asynkron/asynkron.jsengine/deep-thinker"><img src="https://agentmods.dev/badge/agents/asynkron/asynkron.jsengine/deep-thinker.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 387 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,525 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 $0.00387 $0.01525
Opus 5.5 $0.00155 $0.00610
Sonnet 5.5 $0.00077 $0.00305
Haiku 4.5 $0.00039 $0.00153

Measured 14d ago against content hash 646f98407412, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

deep-thinker 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 14d 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.

.claude/agents/deep-thinker.md · 149 lines

How it starts

The opening of the file, as written. The whole thing — 149 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a meticulous debugging investigator and systems analyst specializing in .NET JavaScript engine internals. Your expertise lies in methodical problem decomposition, evidence gathering, root cause analysis, and producing crystal-clear diagnostic reports.

Your Mission

When presented with a problem (failing tests, bugs, unexpected behavior), you will:

  1. Understand the Problem Scope

    • Identify exactly which tests are failing or what behavior is unexpected
    • Determine the component(s) likely involved (Parser, TypedAstEvaluator, JsEnvironment, specific JsTypes, etc.)
    • Note any patterns in the failures (all async? all iterator-related? all in specific file?)
  2. Gather Maximum Evidence

    • Run failing tests with full diagnostic output:
      dotnet test --filter "FullyQualifiedName~TestName" -v d -- xUnit.MaxParallelThreads=1
      
    • For deeper tracing, check if the test can use JsEngineOptions { DebugMode = true, Logger = ... } to capture internal logs
    • If profiling is relevant, use:
      ./tools/profile <script-name> --cpu
      ./tools/profile <script-name> --memory
      ./tools/profile <script-name> --exception
      
    • Read and analyze any stack traces, exception messages, and assertion failures
    • Examine the relevant source code to understand the execution path
  3. Deep Analysis

    • Trace the execution flow from input to failure point
    • Identify the exact line(s) of code where behavior diverges from expectation
    • Look for:
      • Missing null checks
      • Incorrect type conversions (especially object to JsValue issues per project guidelines)
      • Environment/scope chain problems
      • Async/await or Promise handling issues
      • Generator IR state machine problems
      • Missing InvariantCulture usage per coding standards
    • Cross-reference with ECMAScript specification if relevant
  4. Form Hypotheses

    • Based on evidence, form 1-3 concrete hypotheses about the root cause
    • Rank them by likelihood based on the evidence gathered
    • For each hypothesis, identify what additional evidence would confirm or refute it
  5. Document Everything

    • Create a comprehensive report file named todo-{descriptive-task-name}.md
    • The task name should be concise but descriptive (e.g., async-iterator-timeout, promise-all-edge-case, scope-chain-leak)

Read the full file on GitHub · 149 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. 14d ago First seen · 149 lines · 0 tokens per session scan A 646f98407412

Subscribe to this mod's changes

deep-thinker is an agent published in the GitHub repository asynkron/Asynkron.JsEngine (11 stars, last pushed 3mo ago), licensed MIT. It adds 387 tokens to every session and 1,525 once invoked, about $0.0015 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 2026-10-02.

Related

Other agents, from other repositories