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 skills add wrislin1/civ6-mcp --skill civ6-arena-livegit clone --depth 1 https://github.com/wrislin1/civ6-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/skills/wrislin1/civ6-mcp/civ6-arena-live)<a href="https://agentmods.dev/skills/wrislin1/civ6-mcp/civ6-arena-live"><img src="https://agentmods.dev/badge/skills/wrislin1/civ6-mcp/civ6-arena-live/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/wrislin1/civ6-mcp/civ6-arena-live"><img src="https://agentmods.dev/badge/skills/wrislin1/civ6-mcp/civ6-arena-live.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 | $0.00046 | $0.03292 |
| Opus 5 | $0.00023 | $0.01646 |
| Sonnet 5 | $0.00009 | $0.00658 |
| Haiku 4.5 | $0.00005 | $0.00329 |
Grade A, and why
civ6-arena-live 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 11d 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.
2. The gateway is reachable and the local model id is actually served (`curl .../v1/models`) — model names drift (`gemma4:26b`, not `gemma4-26b-cpp`). How it starts
The opening of the file, as written. The whole thing — 250 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Civ6 Arena Live
Overview
Live arena operation has two separate states: the Civilization game turn state and the external watcher process state. Verify both before telling the user to end a turn or before claiming the game is safely back to the human.
Environment
- WSL gaming PC:
[email protected] - WSL repo:
~/projects/civ6-mcp - Native Windows companion checkout:
C:\Users\wrisl\dev\civ6-mcp(/mnt/c/Users/wrisl/dev/civ6-mcpfrom WSL). It is a GitHub clone, not a separate codebase. - Known-good hybrid watchers (both verified live):
- Codex: player 1
local:qwen3-coder:30b, player 2cli-codex:gpt-5.5 - Claude: player 1
local:gemma4:26b, player 2cli-claude:(empty model = Claude default) - both with
--max-puppet-turns 2and--idle-poll-limit 1800
- Codex: player 1
Operating Pattern
When the requested save is not already verified in-game, REQUIRED CONDITIONAL REFERENCE: read references/windows-launcher.md before starting a watcher. It covers the native companion checkout, Windows Application Control, OCR loading, exact-state verification, and FireTuner release.
CLI-provider pre-flight (when a watcher uses cli-claude or cli-codex), on the target host before arming — a failed seat only surfaces mid-handoff otherwise:
- The provider binary is on PATH and authenticated (e.g. a trivial
claude -p/codex execactually returns). A host that has only ever run one provider may not have the other set up. - The gateway is reachable and the local model id is actually served (
curl .../v1/models) — model names drift (gemma4:26b, notgemma4-26b-cpp). .mcp.jsonis present in the repo CWD (project auto-discovery needs it).- Tip: a diagnostic
claude -pcallingget_game_overviewwith no game loaded should return the FireTuner4318connection error — that proves the civ6 tools load on this host.
Preregistration gate (added after v8, 2026-08-30)
For every recorded experiment, resolve and record the actual endpoint for
each configured model/gateway pair before launch; green health alone is not
enough. Confirm that unrelated local-LLM and benchmark jobs use
non-overlapping endpoints. For example, if living-emerald is active, verify
and record that it is actually pinned to home-llm rather than the experiment
gateway; this is an example, not a permanent allocation. Warm every configured
model on its resolved endpoint and record representative response latency.
Block launch until the endpoint allocations are non-overlapping and all of
this evidence is recorded.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- references/windows-launcher.md 2.7 KB
- scripts/arena-live-status.sh 1.9 KB runs code
- scripts/firetuner-owner-map.sh 864 B runs code
- scripts/preflight-tool-calling.py 4.4 KB runs code
- scripts/start-hybrid-watch.sh 8.6 KB runs code
- scripts/stop-arena-watchers.sh 1.7 KB runs code
- scripts/windows-civ6-launcher.sh 645 B runs code
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.
- 11d ago First seen · 250 lines · 46 tokens per session scan A 7b09c424a307
civ6-arena-live is a skill published in the GitHub repository wrislin1/civ6-mcp (0 stars, last pushed 10d ago), licensed MIT. It adds 46 tokens to every session and 3,292 once invoked, about $0.0002 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 skills, from other repositories
gameobject-component-destroy
Destroy one or more Components from a target GameObject. Missing (null) components are skipped — they cannot be destroyed. Use 'gameobject-find' and 'gameobject-component-get' to identify the components first.
unity-version-split
Split a C# file into Unity 6.5+ and pre-Unity 6.5 variants. Use when a file needs different implementations for different Unity versions due to API changes (e.g., EntityId vs int, GetEntityId vs GetInstanceID).
godot-signals-groups
Build event-driven, decoupled Godot 4.7 gameplay with signals and node groups: declare and emit custom signals, connect with Callables (incl. bind/one-shot), and broadcast to many nodes via groups and callgroup. Use when wiring node communication in a Godot project, replacing tight references with signals…
unity-addressables
Manage Addressables groups, entries, profiles and content builds (com.unity.addressables, reflection-based).
motion
How an agent turns a character mesh into a usable animated FBX — and how to judge whether the result is shippable.
threejs-exposure-color-grading
Build a measured exposure and grading path in Three.js. Use for a 64x36 encoded luminance meter, asynchronous readback, weighted log-average exposure, asymmetric adaptation, single tone-map ownership, and a generated 32-cube post-tone-map LUT.