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/khamel83/argus/agents-mdgit clone --depth 1 https://github.com/Khamel83/argusWhat 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 | $0.03021 | $0.03021 |
| Opus 5 | $0.01510 | $0.01510 |
| Sonnet 5 | $0.00604 | $0.00604 |
| Haiku 4.5 | $0.00302 | $0.00302 |
Grade A, and why
argus AGENTS.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 — 240 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Argus
What this file is for: the canonical guide for AI coding agents working in this repository (Claude Code, Codex, Cursor, Copilot, OpenCode, and friends). Human contributors should start with CONTRIBUTING.md; background on the project lives in CONTEXT.md and README.md. CLAUDE.md is a short pointer back here with Claude-Code-specific notes.
Overview
Search infrastructure for AI agents: 14 providers, topology-aware routing, WolframAlpha computed answers, and a 12-step content extraction chain. Provider adapters: SearXNG (self-hosted, aggregates 70+ engines, disabled by default), DuckDuckGo, Yahoo (scraped), GitHub, WolframAlpha (free API, computed answers), Brave, Tavily, Exa, Linkup, Parallel AI (monthly credit for eligible accounts with a card on file), Serper, You.com, Valyu, SearchAPI. Tier-based routing: free providers first, monthly recurring next, one-time credits last. Budget enforcement skips exhausted providers automatically. Multi-turn sessions use the configured SQL repository.
Features
- Topology-aware acquisition — Argus knows if it's on a residential IP or datacenter, routing search and extraction automatically to avoid blocks and minimize network hops.
- Adaptive Domain Memory — Learns which domains fail from datacenter IPs but succeed from residential ones, automatically routing future requests for those domains to residential egress.
- Universal Provenance — Every search result and extraction is tagged with
egress(residential|datacenter),machine, andsource_type. - Intelligent Routing — Tier-based routing: free providers first, monthly recurring next, one-time credits last. Budget enforcement skips exhausted providers automatically.
- 12-step Extraction — Trafilatura → Crawl4AI → Playwright → Jina → Valyu → Firecrawl → You → Archive with quality gates and completeness assessment.
Two Deployment Tiers
Tier 1: No server (API keys only)
pip install argus-search— works immediately with DuckDuckGo + Yahoo- Add WOLFRAM_APP_ID for computed answers (math, facts, conversions — 2,000 free/month)
- Add API keys for 7,000+ more free monthly queries; eligible Parallel accounts with a card on file can bring the total to as much as 12,000+
- Extraction via external APIs only (Jina, Valyu Contents, Firecrawl, You.com Contents, Wayback)
- Default SQLite persistence (no database server required)
- Works on any machine with Python 3.11+ (laptop, Mac Mini, Pi, cloud VM)
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 · 240 lines · 3,021 tokens per session scan A 6f197f54aa7d
argus AGENTS.md is an instructions file published in the GitHub repository Khamel83/argus (4 stars, last pushed 2d ago), licensed MIT. It adds 3,021 tokens to every session, about $0.0151 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.
Other instructions, from other repositories
apple-browsers AGENTS.md
AGENTS.md instructions for duckduckgo/apple-browsers, covering mandatory rules and opening a pr with snapshot changes.
web-researcher-mcp CLAUDE.md
Instructions for zoharbabin/web-researcher-mcp, covering claude.md — web-researcher-mcp, commands, contribution rules, architecture and non-obvious top-level dirs.
apple-browsers pixels.instructions.md
Instructions for duckduckgo/apple-browsers, a project described as: DuckDuckGo iOS & macOS browsers.
web-search-mcp CLAUDE.md
Instructions for sydasif/web-search-mcp, covering claude.md, common commands, development, testing and architecture.
cc-websearch CLAUDE.md
Instructions for Djarvur/cc-websearch, covering project, constraints, technology stack, what not to add and conventions.
agent-skills AGENTS.md
Instructions for youdotcom-oss/agent-skills, covering agent instructions, tooling discovery, minimal-implementation directive, style enforcement and workflow.