config
Configure prompt-cache warming (keep-alive pings for idle Claude Code sessions) per scope - global, CLI, desktop app, project folder or single session -…
Skill by FiredMosquito831
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/firedmosquito831-config
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: data, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/config
Tools it pre-approves 1
From allowed-tools in SKILL.md: the agent may use these without asking you each time.
Bash(node:*)
What's inside
- Skill name
config- Description
- Configure prompt-cache warming (keep-alive pings for idle Claude Code sessions) per scope - global, CLI, desktop app, project folder or single session - interval, idle cap, when to warm, dashboard. Use when the user asks to change how often the cache is warmed, to stop or start cache warming somewhere, or to open the cache-warm dashboard.
- License
- MIT
- Allowed tools
- Bash(node:*)
- Instructions
- 28 lines in SKILL.md · view on GitHub
Before installing
- Pre-approves some shell commands
Agent Skills spec issues
- name: missing
Files (1)
- SKILL.md
SKILL.md (first lines)
Control the cache-warm plugin through its CLI: `node "${CLAUDE_PLUGIN_ROOT}/scripts/ccw.mjs" <command> [scope flag]`.
| User wants | Command |
| :--- | :--- |
| See effective settings and where each comes from | `config` |
| Change ping interval | `interval <minutes>` or `interval auto` |
| Stop pinging after N idle minutes | `set maxIdleMinutes <N>` (0 = never) |
| Warm only while background work is pending (default), or any idle session | `when background-work` / `when always` |
| Skip small sessions | `set minContextTokens <N>` |
| On / off | `on` / `off` |
| Go back to inheriting one setting / everything at a scope | `unset <key>` / `reset` |
| List every scope that has settings | `scopes` |
| Open the dashboard | `dashboard` |
| Something seems broken | `doctor`, then `events 30` |
Scope flags (default inside a session: this session's surface, CLI or desktop app, so one never changes the other; `--global` for both): `--session` (this session), `--project` (the current folder and its subfolders; `--project=<folder>` for another), `--cli`, `--desktop`, `--ide`. Most specific wins: session > project > surface > global > plugin tab > defaults. Change only the scope the user asked for. Use `--global` only when the user explicitly asks for every session (CLI and desktop app); never to fix one session or one surface.
Changes apply live; never tell the user to restart. Never create scheduled tasks (CronCreate) for cache warming: the plugin's Stop hook times the pings and resets the timer on every request.
## Rules of thumb to share when asked
- The interval must be shorter than the cache TTL (5 minutes by default on API keys, 1 hour on a Claude subscription or with `CLAUDE_CODE_PROMPT_CACHE_TTL=1h`). `auto` handles this: 50 minutes on a 1-hour cache.
- The timer counts from the conversation's last request, so a session in use is never pinged.
- A ping costs one cache read of the whole context; a cold return costs one cache write of it. `status` prints both.
Arguments: $ARGUMENTSFrom FiredMosquito831/claude-cache-warm, MIT.
More from FiredMosquito831/claude-cache-warm
- 0.549score
on
FiredMosquito831
Turn prompt-cache warming on (globally, or for this session / project / desktop / cli)
SkillAFreshness: AClaude Code +9 - 0.549score
off
FiredMosquito831
Turn prompt-cache warming off (globally, or for this session / project / desktop / cli)
SkillAFreshness: AClaude Code +9 - 0.527score
dashboard
FiredMosquito831
Start the cache-warm web dashboard and open it in the browser (live sessions, global / surface / project / session settings, cache analytics)
SkillAFreshness: AClaude Code +9 - 0.524score
status
FiredMosquito831
Show prompt-cache warming status, next ping countdown and per-ping cost for every live session
SkillAFreshness: AClaude Code +9 - 0.522score
test
FiredMosquito831
Self-test prompt-cache warming in this session (CLI or desktop app) - one keep-alive ping a minute after this turn, verified against the transcript
SkillAFreshness: AClaude Code +9 - 0.512score
statusline
FiredMosquito831
Add (or remove) the cache-warm segment in the Claude Code status line - warm on/off, context tokens, cached tokens, time until the cache expires
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 8, 2026 (20 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Website
- cachewarm.danubelabs.net
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from FiredMosquito831
- 0.557score
kb-integrity
FiredMosquito831
Load when a vellum check failed, or when muse routes integrity work.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
cold-read
FiredMosquito831
Load for /vellum:cold-read, after every major editorial program, and before export.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
gates
FiredMosquito831
Load when muse accepts a pivotal chapter, before export, or for any scoring task.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
demolition
FiredMosquito831
Load when the author invokes demolition on a chapter draft, or when muse offers it before marking a chapter final.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
qi-layer
FiredMosquito831
Use when writing or maintaining AGENTS.md, .context/CONTEXT.md, or CLAUDE.md mirrors: keep intent docs minimal and load-bearing.
SkillBFreshness: B★ 2Claude Code +9 - 0.553score
export
FiredMosquito831
Load for /vellum:export.
SkillBFreshness: B★ 2Claude Code +9
Cite this page
config skill. Agent Mart. https://agentmart.dev/skills/firedmosquito831-config (last updated 2026-10-24)