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.
npx agentmods add instructions/umage-ai/codeshellmanager/claude-mdgit clone --depth 1 https://github.com/umage-ai/CodeShellManagerWhat 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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.09292 | $0.09292 |
| Opus 5 | $0.04646 | $0.04646 |
| Sonnet 5 | $0.01858 | $0.01858 |
| Haiku 4.5 | $0.00929 | $0.00929 |
Grade A, and why
CodeShellManager CLAUDE.md 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.
How it starts
The opening of the file, as written. The whole thing — 442 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CodeShellManager
A WPF .NET 10 multi-terminal host for Claude Code (and other CLI tools). Runs multiple pseudo-terminal sessions in a tabbed/grid layout with full-text search, alert detection, git status, and session persistence.
Build & Run
# From repo root
dotnet build src/CodeShellManager/CodeShellManager.csproj
dotnet run --project src/CodeShellManager/CodeShellManager.csproj
# Or open CodeShellManager.slnx in Visual Studio / Rider
Requirements: .NET 10 SDK, Windows 10/11 (uses ConPTY + WebView2)
Visual Studio: Hot Reload disabled
Properties/launchSettings.json ships with "hotReloadEnabled": false, and the csproj sets <MetadataUpdaterSupport>false</MetadataUpdaterSupport> in Debug. Both are workarounds for a System.ExecutionEngineException that crashes the app on F5 under .NET 10.0.8 + VS 18 — Microsoft.Extensions.DotNetDeltaApplier.dll faults during its own startup before any managed code runs. Ctrl+F5 (Start Without Debugging) is unaffected either way. Remove both when the runtime bug is fixed upstream.
Command-line flags
| Flag | Effect |
|---|---|
--clean |
Debug isolation mode — see below. |
--clean (parsed in App.OnStartup, exposed as App.CleanStart):
MainWindow.OnLoadedskips the restore loop and clears the in-memorySessionManager— both sessions AND groups — so any new work in the run starts from a blank slate.MainViewModel.SaveStateAsyncshort-circuits — nothing is written tostate.jsonfor the entire run. Window bounds, layout changes, settings tweaks, and any sessions created during the clean run are all discarded on exit.- The user's prior
state.jsonsurvives the run untouched, so this is the safe way to test from a blank slate.
Architecture
Key layers
PTY (ConPTY) → PseudoTerminal → TerminalBridge → WebView2 (xterm.js)
↓
OutputIndexer → SQLite FTS5
AlertDetector → SessionViewModel.RaiseAlert()
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.
- 2d ago First seen · 442 lines · 9,292 tokens per session scan A 1f616f936bc3
CodeShellManager CLAUDE.md is an instructions file published in the GitHub repository umage-ai/CodeShellManager (21 stars, last pushed 3d ago), licensed MIT. It adds 9,292 tokens to every session, about $0.0465 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.
Other instructions, from other repositories
DevoxxGenieIDEAPlugin GEMINI.md
Instructions for devoxx/DevoxxGenieIDEAPlugin, a project described as: DevoxxGenie is an agentic plugin for IntelliJ IDEA that uses local LLM's (Ollama, LMStudio, GPT4All, Jan and Llama.cpp) and Cloud based LLMs to help review, test, explain your project code. Latest version now also supports Spec Driven Development…
packmind packmind-frontend-error-management.instructions.md
Instructions for PackmindHub/packmind: A global error boundary is already configured at the root level (in root.tsx) that automatically catches all page crashes, 404s, and loader failures. This standard defines when and how to use explicit... .
Ctxo CLAUDE.md
Instructions for alperhankendi/Ctxo, covering claude.md — ctxo, project overview, quick reference, dev (pnpm workspace) and usage (consumer).
apex-accelerator no-interactive-shell.instructions.md
Prevents interactive shell prompts and long-output terminal replays from being injected into chat. Forbids -i flags on mv/rm/cp, read -p, and confirm prompts (incl. inside bash -c '...'). Pipe long output to files. Scoped to chat-context-loaded files; skill references/ and templates/ are exempt because they hold…
apex-accelerator iac-bicep-best-practices.instructions.md
Bicep-specific IaC best practices for Azure templates. Security baseline, naming, AVM mandate, anti-patterns.
llm-ide-rules pytest-integration-tests.instructions.md
Instructions for iloveitaly/llm-ide-rules, covering pytest integration tests and example integration test.