Borrowing it
Nothing to install: this file belongs to rakaarwaky/qwen-web-arwaky. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/rakaarwaky/qwen-web-arwaky/main/AGENTS.mdgit clone --depth 1 https://github.com/rakaarwaky/qwen-web-arwakyWrote 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/instructions/rakaarwaky/qwen-web-arwaky/agents-md)<a href="https://agentmods.dev/instructions/rakaarwaky/qwen-web-arwaky/agents-md"><img src="https://agentmods.dev/badge/instructions/rakaarwaky/qwen-web-arwaky/agents-md/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/instructions/rakaarwaky/qwen-web-arwaky/agents-md"><img src="https://agentmods.dev/badge/instructions/rakaarwaky/qwen-web-arwaky/agents-md.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 | $0.01998 | $0.01998 |
| Opus 5.5 | $0.00799 | $0.00799 |
| Sonnet 5.5 | $0.00400 | $0.00400 |
| Haiku 4.5 | $0.00200 | $0.00200 |
Grade A, and why
qwen-web-arwaky 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 5d 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 — 109 lines — stays where its author put it; the contents beside it link to each section on GitHub.
qwen-web-arwaky — Agent Operating Guide
Agent Overview
Production-grade Python CLI + MCP server that automates chat.qwen.ai with no API key: it sends prompts, waits for the response, extracts it, and saves it locally. Built on the AES 7-layer architecture, enforced in CI by lint-arwaky-cli at zero violations. Before generating code, docs, or suggestions, an agent must know: INPUT is prompt text or -i <prompt>.md, optionally -a <attachment>, driven through an authenticated chat.qwen.ai browser session; OUTPUT is a Markdown file at -o <path> whose extraction relies on pinned DOM selectors the tests lock as regressions, and whose .meta.json sidecar lives under XDG state (~/.local/state/qwen-web-arwaky/meta/), never in the output folder. The CLI is qwen-web-arwaky (alias qwa); the legacy qwen-web-cli and qwc entry points were removed in v6.0.0. Pipeline shape: CLI/MCP surface → agent orchestrator → capabilities → Playwright browser adapter → Markdown output (input → orchestrate → execute → drive → persist), wired once by the root containers. Response style: concise, technical, direct — lead with the command or the diff, explanation after if needed.
Runtime
| Aspect | Detail |
|---|---|
| Language | Python >= 3.10; CI exercises 3.13 (the test matrix entry) |
| Environment | uv-managed, locked by uv.lock; uv lock --check is a CI gate, so a dependency change regenerates the lock in the same commit |
| Artifacts | Build output lands in dist/; generated responses go to the -o/--output path the user names. Never use /tmp for build output a reviewer must find |
| Package Manager | uv |
| Version Probe | python --version |
| Setup | export UV_PROJECT_ENVIRONMENT="$PWD/.venv" && uv sync && uv run python -m playwright install chromium |
| Activation | Run every command through uv run; the venv is isolated from the working tree |
Project Structure
.
├── modules/shared/src/ # taxonomy_*, contract_*, utility_*
├── modules/browser/src/ # agent_browser_orchestrator, capabilities_browser_*
├── modules/config/src/ # agent_config_orchestrator, capabilities_config_*
├── modules/jobs/src/ # agent_job_orchestrator, capabilities_folder_compiler, capabilities_job_storage
├── modules/logging/src/ # agent_logging_orchestrator, capabilities_metrics_counter, capabilities_observability_setup
├── modules/prompt/src/ # agent_prompt_orchestrator, agent_shared_flow_orchestrator, capabilities_*
├── modules/prompt_runtime/src/ # agent_prompt_runtime_orchestrator, capabilities_* (browser/stream/persist layer)
├── modules/session/src/ # agent_session_orchestrator, capabilities_session_*
├── modules/swarm/src/ # agent_swarm_orchestrator, capabilities_swarm_runner
├── modules/update/src/ # agent_update_orchestrator, capabilities_update_manager
├── modules/cli/src/ # surface_cli_*, root_cli_container
├── modules/tui/src/ # surface_tui_* (Textual TUI layer; depends on cli, never vice versa)
├── modules/mcp/src/ # surface_mcp_*, root_mcp_container
├── modules/shared/templates/ # role templates (package data of the shared member), free of HTML-escaped tokens
├── modules/root_cli_main_entry.py # CLI composition root entry point
├── modules/root_mcp_main_entry.py # MCP composition root entry point
├── modules/root_core_container.py # shared dependency wiring used by the CLI and MCP roots
├── tests/ # shared fixtures (conftest, pipeline_fixtures, helpers) — not a pytest root; pytest lives in modules/*/tests/
├── benches/ # benchmark suite, marker: benchmark
├── lint_arwaky.config.yaml # architecture rules enforced by lint-arwaky-cli
├── pyproject.toml # build config, entry points, and pinned dependencies
├── uv.lock # lockfile; `uv lock --check` is a CI gate
└── Containerfile # containerized execution, driven by scripts/podman.sh
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.
- 5d ago Changed · -32 lines · +388 tokens per session 8ce71acfbeb0
- 16d ago Changed · +2 lines · +35 tokens per session 37ab8408ef7e
- 20d ago Changed · -9 tokens per session 7f4affe7576f
- 23d ago Changed · +4 tokens per session 0c0094a7de58
- 25d ago Changed · +1 lines · +38 tokens per session bd87c0595e04
- 26d ago Changed · -1 tokens per session 96d58d4cc3fc
- 28d ago Changed · -201 lines · -1,743 tokens per session scan B → A 9036e3eb5298
- 1mo ago Changed · +81 lines · +520 tokens per session 61f47bb1f159
- 1mo ago Changed · +10 lines · +162 tokens per session 516181b89d4d
- 1mo ago First seen · 248 lines · 2,604 tokens per session scan B 3eda24a8e0a2
qwen-web-arwaky AGENTS.md is an instructions file published in the GitHub repository rakaarwaky/qwen-web-arwaky (0 stars, last pushed yesterday), licensed MIT. It adds 1,998 tokens to every session, about $0.0080 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.
qwen-web-arwaky AGENTS.md compared with instructions from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| qwen-web-arwaky AGENTS.mdrakaarwaky/qwen-web-arwaky | Instructions file | 0 | yesterday | MIT | A | 1,998 | Codex, OpenCode |
| vscode buildNext.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 9,719 | GitHub Copilot |
| next.js AGENTS.mdvercel/next.js | Instructions file | 143k | yesterday | MIT | A | 7,461 | Codex, OpenCode |
| langchain AGENTS.mdlangchain-ai/langchain | Instructions file | 149k | today | MIT | A | 4,680 | Codex, OpenCode |
| vscode oss-third-party-notices.instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 5,516 | GitHub Copilot |
| vscode copilot-instructions.mdmicrosoft/vscode | Instructions file | 195k | today | MIT | A | 3,093 | GitHub Copilot |
| deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness | Instructions file | 275k | 2d ago | MIT | A | 4,313 | Codex, OpenCode |