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/franklinnolasco7/quickshell-mcp/agents-mdgit clone --depth 1 https://github.com/franklinnolasco7/quickshell-mcpWrote 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/franklinnolasco7/quickshell-mcp/agents-md)<a href="https://agentmods.dev/instructions/franklinnolasco7/quickshell-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/franklinnolasco7/quickshell-mcp/agents-md.svg" alt="Measured on agentmods" 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 | $0.04365 | $0.04365 |
| Opus 5 | $0.02183 | $0.02183 |
| Sonnet 5 | $0.00873 | $0.00873 |
| Haiku 4.5 | $0.00436 | $0.00436 |
Grade A, and why
quickshell-mcp AGENTS.md scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
rtk curl -sL https://quickshell.org/about/ -o tests/fixtures/about.html How it starts
The opening of the file, as written. The whole thing — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Guidance for AI agents (Claude Code, opencode, etc.) working on this repository.
Quick start
pip install -e '.[dev]'for the venv (ornix developfor the CI-equivalent shell)pytest -qfor the offline suite; see Build/test commands below for the full list
Scope
Governs agent behavior within quickshell-mcp only. This is a single-author personal tool, so keep changes proportional: no release automation, no features without a stated consumer, no dependencies without a reason.
Project overview
An MCP server that serves live Quickshell documentation from quickshell.org so LLMs read real docs instead of hallucinating API surface (the API has shifted heavily between releases). Three companion sources round it out: Qt type docs from doc.qt.io (base types like Rectangle that Quickshell configs import but quickshell.org doesn't document), the official quickshell-examples repo via its Gitea API, and real-world QML shells (Caelestia, Noctalia, end-4's dots-hyprland) indexed from their GitHub repos. Everything is fetched at query time; nothing is hardcoded.
Project structure
quickshell_mcp/, the package itself:config.py: URLs, repo identities, strip-selector lists, TTL/retry constantscaches.py: 30-minute in-memory cache per fetch, plus a disk layer that persists built indexes (guide/type/qt) across restarts (30-day default TTL);QUICKSHELL_DOCS_MCP_DISK_CACHEenv var to disable/redirectutils.py: logging, HTTP client,_fetch_raw(retry + stats), citation prefix; domain modules callutils._fetch_rawby module attribute so tests monkeypatch one seamversions.py,extraction.py: runtime version discovery; HTML→Markdown with per-source strip rulessources/: shared-services layer (data-source access, index building, parsing, project context, orchestration). No MCP decorators here, and it never importscapabilities/(dependencies point down). Roughly grouped as: doc/content sources (docs,qt_docs,examples,implementations), code-shaping tools (compat,migrate,validate,generate,refactor), project-facing tools (assistant,explain_error,find_pattern,search_all,project,project_intel,project_validate,project_memory), and runtime/analysis tooling (intelligence,agent_tools,knowledge2,ecosystem,perf_profile,runtime_profile,runtime_session,runtime_debug,runtime_test,runtime_visual,ui_runtime,style_match,adapters/). See individual module docstrings for specifics.capabilities/: capability/domain layer between the tools and the sources, one module per capability (knowledge,knowledge2,validation,generation,migration,project,runtime,inspection,testing,performance,debugging,adapters,ecosystem,intelligence,agent,assistant). Each declaresCAPABILITY_NAME/CAPABILITY_TOOLS/CAPABILITY_DEPENDS_ONand re-exports the source entry points;registry.pyaggregates the tool→capability map, a cycle-checkeddependency_order(), and the safety model. Every capability carries asafety_level(read-onlydefault;runtimeandtestingaremutating), with per-tool overrides in_TOOL_SAFETY_LEVELS(apply_patch,ui_set_property,ui_invoke,reloadaremutating;ui_evalishigh-risk).server.py: FastMCP instance, every@mcp.tool()wrapper + docstring (the trigger surface),main(), and re-exports of helper names so tests keep addressingsrv.<helper>; tool logic is routed throughcapabilities/
tests/: offline pytest suite driven by saved snapshots intests/fixtures/.conftest.pyprovidesmock_fetch(patchesutils._fetch_raw),docs_fixture_urls,http_404, and an autouse cache/stats reset.scripts/:smoke_test.py(end-to-end stdio client), standalone headless CI entrypoints (ci_validate.py,ci_screenshot.py,ci_runtime_test.py,ci_api_compat.py,ci_migration.py), and QA tooling (quality_audit.py,benchmark.py). Each CI script exposes a testablerun()function separate from its argparsemain().flake.nix/Dockerfile/.github/workflows/ci.yml+.releaserc.js: packaging + CI + semantic release. The flake has acidevShell (nix develop .#ci) withquickshell,weston,grim, andimagemagickfor headless runtime/screenshot jobs.CLAUDE.mdis a symlink to this file.
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 First seen · 180 lines · 4,365 tokens per session scan A e97861a8d5bc
quickshell-mcp AGENTS.md is an instructions file published in the GitHub repository franklinnolasco7/quickshell-mcp (0 stars, last pushed 6d ago), licensed MIT. It adds 4,365 tokens to every session, about $0.0218 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
openrouter-mcp-multimodal AGENTS.md
AGENTS.md instructions for stabgan/openrouter-mcp-multimodal, covering agent instructions, before you ship, releasing (read this before publishing), short version and version files (must all match package.json).
Plonk AGENTS.md
AGENTS.md instructions for ostapondo/Plonk, covering agent rules, layout, adding a module, build & verify and code style.
cyxj-groksearch CLAUDE.md
Claude Code instructions for chenyuxiaojin/cyxj-groksearch, covering claude.md, 常用命令 and 结构与坑.
mcp-dockhand copilot-review.instructions.md
Copilot instructions for strausmann/mcp-dockhand: When reviewing pull requests in this repository.
minecraft-dev-mcp AGENTS.md
Instructions for MCDxAI/minecraft-dev-mcp, covering agent handbook – minecraft dev mcp, project snapshot, what agents should prioritize, architecture wayfinder (src/) and available mcp tools (for llm surfaces).
bray AGENTS.md
AGENTS.md instructions for forgesworn/bray, covering agents.md — nostr-bray, what this is, build & test, architecture and tool groups.