Argus minimal-engineering.instructions.md

A set of instructions for keeping an agent-based software project as small and direct as possible while still checking real boundaries such as input, permissions, saved data, security, and irreversible actions.

In plain words
What is it for?
Use it when designing or reviewing agent workflows, orchestration, validation steps, ownership, and escalation rules.
Why use it?
It reduces unnecessary roles, repeated checks, wrappers, retries, and artifacts that make a system harder to understand without improving its decisions.

Instructions file for GitHub Copilot

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/lbx154/argus/minimal-engineering
Clone the repo
git clone --depth 1 https://github.com/lbx154/Argus

Made for: GitHub Copilot.

Per session 616 This file is loaded in full into every session.
When invoked 616 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.

ModelPer sessionOnce invoked
Fable 5 $0.00616 $0.00616
Opus 5 $0.00308 $0.00308
Sonnet 5 $0.00123 $0.00123
Haiku 4.5 $0.00062 $0.00062

Measured 2d ago against content hash 7a2553af69aa, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

Argus minimal-engineering.instructions.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.

.github/instructions/minimal-engineering.instructions.md · 52 lines

What it actually says

Minimal engineering

  • Use the shortest control path that satisfies the current requirement. Add a role, guard, retry, wrapper, or artifact only when it resolves a reachable uncertainty and its failure changes the next action.
  • Validate real input, authority, persistence, security, and irreversible boundaries. Trust established internal invariants instead of repeating the same check in every layer.
  • Keep one owner for one coherent deliverable. Split work only for a hard dependency, independent information source, or independent authority such as Reviewer acceptance.
  • Pass one canonical contract plus role-specific deltas. Do not restate the operator request, plan, checkpoint, and review policy in every role prompt.
  • Treat an accepted decisive check as settled until its inputs change or a contradiction appears. Do not add another validation-only task or rerun an unchanged check for ceremony.
  • Repair routine local failures locally. Escalate only when scope, semantics, integrity, authority, or irreversible effects change.
  • Do not introduce hashes, including SHA-256, for identifiers, migrations, freshness, or correctness. Prefer direct names and explicit persisted state; do not add defensive fallbacks or guards for states the current contract cannot reach.
  • Without a demonstrated current requirement, do not add UUIDs, random tokens, custom idempotency/deduplication keys, retries, backoff, circuit breakers, fallback chains, compatibility layers, speculative locks, future fields, or placeholder interfaces. Existing framework and upstream invariants are the default; validate only real external-input, permission, persistence, security, and irreversible-action boundaries.
  • Keep errors transparent. Do not turn defects into empty results, default values, broad catches, or success-shaped degradation.
  • Do not create helpers, wrappers, services, factories, or abstractions unless they remove existing semantic duplication, enforce a real boundary, or make the current call path materially easier to understand. Prefer direct local code and stop when the requested behavior is complete.
  • Parallel and Team execution must be real rather than ceremonial: independent work, disjoint writable paths, explicit completion evidence, one lead-owned synthesis, and normal Reviewer acceptance. DSH native subagents may assist flexibly inside that structure; do not confuse them with or forbid the Argus Team mechanism the operator requested.
  • Test the changed surface and reachable consumers. Run an end-to-end smoke test for a newly exercised path; broaden only when the blast radius is broad.

For example, coupled files produced by one Engineer and checked by one Reviewer are one task, not an implementation task followed by a second review task. Conversely, credential handling and publication remain explicit authority boundaries even when bypassing them would be shorter.

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. 2d ago First seen · 52 lines · 616 tokens per session scan A 7a2553af69aa

Subscribe to this mod's changes

Argus minimal-engineering.instructions.md is an instructions file published in the GitHub repository lbx154/Argus (293 stars, last pushed 2d ago), licensed MIT. It adds 616 tokens to every session, about $0.0031 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.

Related

Other instructions, from other repositories

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

buildNext

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

next.js AGENTS.md

Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).

microsoft/vscode · 5,001 tokens

spec-kit AGENTS.md

Instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.

github/spec-kit · 7,040 tokens

langchain AGENTS.md

Instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,345 tokens