What is pskit CLAUDE.md?

Repository instructions for PSKit, a Python MCP server that exposes PowerShell tools. PowerShell is Microsoft’s command-line and scripting environment.

Use them when changing PSKit’s server, session manager, safety pipeline, neural command scorer, PowerShell tools, or tests.

They give the coding agent the project’s architecture, commands, safety rules, and implementation constraints in one place.

pskit CLAUDE.md — Instructions file by Nickalus12

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 instructions/nickalus12/pskit/claude-md

Clone the repo

git clone --depth 1 https://github.com/Nickalus12/pskit

Per session 748 This file is loaded in full into every session.

When invoked 748 The same file — it is already loaded in full.

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.

Model	Per session	Once invoked
Fable 5	$0.00748	$0.00748
Opus 5	$0.00374	$0.00374
Sonnet 5	$0.00150	$0.00150
Haiku 4.5	$0.00075	$0.00075

Measured yesterday against content hash b24ca61f4e3b, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

pskit 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 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.

CLAUDE.md · 64 lines

How it starts

The opening of the file, as written. The whole thing — 64 lines — stays where its author put it; the contents beside it link to each section on GitHub.

PSKit — Developer Context

PSKit is a standalone Python MCP server exposing 19 PowerShell tools with a 5-tier neural safety pipeline. Used by Loom as its PowerShell execution layer.

Architecture

File	Purpose
`src/pskit/server.py`	FastMCP server — 19 tools, 4 prompts, 1 resource
`src/pskit/manager.py`	PSKitManager — session pool, 5-tier safety pipeline, result cache
`src/pskit/kan_engine.py`	PSKitKANEngine — 24-feature KAN neural command scorer
`src/pskit/kan.py`	Vendored EfficientKAN (MIT) — pure PyTorch KAN implementation
`src/pskit/tools/PSKitTools.psm1`	19 PowerShell functions loaded into every session
`tests/test_manager.py`	32 unit tests (no live PS session required)

Key Invariants

All PS execution flows through PSKitManager.execute(script) — never bypass
Safety pipeline lives in _execute_inner() — 5 tiers run in order, never skip
Result cache invalidates on every write command success (conservative but correct)
Session pool: 3 named-pipe connections to pwsh, checkout/checkin pattern
KAN scorer: PSKitKANEngine.score_command(cmd) → float 0.0 (safe) – 1.0 (dangerous)

Development Commands

python -m pytest tests/ -q          # run all 32 unit tests
ruff check src/                      # lint
python -m pskit                      # start MCP server on stdio

Adding a New Tool

Add PS function to src/pskit/tools/PSKitTools.psm1
Add function name to Export-ModuleMember at the bottom of the psm1
Add @mcp.tool() in server.py calling _require().execute("New-PSKitFunction ...")
Write tests in tests/test_manager.py (mock the PS session — don't need real pwsh)

Adding a New Prompt

@mcp.prompt()
def my_workflow() -> str:
    """One-line description shown in Claude's prompt library."""
    return "Step-by-step instructions telling Claude how to use pskit tools..."

Safety Pipeline (do not modify without understanding all 5 tiers)

Read the full file on GitHub · 64 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.

yesterday First seen · 64 lines · 748 tokens per session scan A b24ca61f4e3b

Subscribe to this mod's changes

pskit CLAUDE.md is an instructions file published in the GitHub repository Nickalus12/pskit (0 stars, last pushed 5d ago), licensed MIT. It adds 748 tokens to every session, about $0.0037 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.

Languages Planning Testing

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

pskit CLAUDE.md