Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/magpie-lang/magpienpx agentmods add skills/magpie-lang/magpie/magpieWrote 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.
[](https://agentmods.dev/skills/magpie-lang/magpie/magpie)<a href="https://agentmods.dev/skills/magpie-lang/magpie/magpie"><img src="https://agentmods.dev/badge/skills/magpie-lang/magpie/magpie/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.
<a href="https://agentmods.dev/skills/magpie-lang/magpie/magpie"><img src="https://agentmods.dev/badge/skills/magpie-lang/magpie/magpie.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 5.1 | $0.00039 | $0.17528 |
| Opus 5.5 | $0.00016 | $0.07011 |
| Sonnet 5.5 | $0.00008 | $0.03506 |
| Haiku 4.5 | $0.00004 | $0.01753 |
Grade A, and why
magpie-engineer 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 2mo 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 — 1,923 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Magpie Language Programming Guide
Use this skill whenever you write, debug, or review Magpie .mp programs.
0) Non-negotiable Rules
- Binary-first: treat compiler diagnostics as ground truth.
- Run
magpie explain <CODE>for any error code you don't recognize. - Use
--output jsonfor machine-readable diagnostics.
- Run
- Diagnose by error code first, then apply the smallest fix.
- Smallest reproducer first -- one tiny
.mpfile, then grow. - Change one dimension at a time (syntax, then types, then ownership, then backend).
- Prove fixes with commands and exit codes.
1) What is Magpie
Magpie is a compiled, SSA-based language with explicit ownership, ARC memory management, and first-class GPU support. Programs are written in a low-level IR-like textual format (CSNF).
Key characteristics:
- SSA form -- every value is defined once, used by name (
%x,%y) - Explicit ownership -- you manually
borrow.shared,borrow.mut,share,clone.shared - ARC-managed heap -- the compiler inserts reference counting; you manage ownership transitions
- Block-structured -- functions contain labeled basic blocks (
bb0:,bb1:) with explicit terminators - No implicit conversions -- types must match exactly;
constsuffixes must match declared types - 5 GPU backends -- SPIR-V, Metal (MSL), PTX (CUDA), HIP (AMD), WGSL (WebGPU)
How Magpie differs from familiar languages:
- There are no expressions or statements -- only SSA instructions and terminators
- Control flow uses
br,cbr,switch-- notif/else/for/while - Variables are SSA locals (
%name) -- assigned exactly once - Functions are
@name-- always prefixed with@ - Types are
TName-- user-defined types start withT - Comments use
;(not//or#); doc comments use;;
2) Module Structure
Every .mp file must have a strict header in exactly this order:
module <module.path>
exports { <exported symbols> }
imports { <import groups> }
digest "<hex string>"
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.
- 2mo ago First seen · 1,923 lines · 39 tokens per session scan A 444b560cf044
magpie-engineer is a skill published in the GitHub repository magpie-lang/magpie (60 stars, last pushed 8mo ago), licensed MIT. It adds 39 tokens to every session and 17,528 once invoked, about $0.0002 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 2026-08-30.
magpie-engineer compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| magpie-engineermagpie-lang/magpie | Skill | 60 | 8mo ago | MIT | A | 39 | Claude Code, Codex |
| omh-rustrlaope/oh-my-hermes | Skill | 5.2k | today | MIT | A | 74 | Claude Code, Codex |
| erlang-otp-behaviorsbenchflow-ai/skillsbench | Skill | 1.9k | 3d ago | Apache-2.0 | A | 52 | Claude Code, Codex |
| axiom-concurrencyCharlesWiltgen/Axiom | Skill | 1.3k | yesterday | MIT | A | 49 | Claude Code |
| tddcodewithmukesh/dotnet-claude-kit | Skill | 807 | 3d ago | MIT | A | 81 | Claude Code |
| splitting-oversized-modulesPostHog/posthog | Skill | 41k | today | none | A | 197 | Claude Code, Codex |
| precisemicroprediction/precise | Skill | 338 | 6d ago | MIT | A | 76 | Claude Code, Codex |