agentleFS
BariBariGoodBariBariGood/manzanas/.agents/skills/run-the-daemon/SKILL.md

run-the-daemon

Skill22 starsChanged 3 months agoApache-2.0View on GitHub

Run or restart the manzanasd daemon - flags, warm pool and safety gates, journal settings, 503 overloaded handling, and install/upgrade via deploy/install.sh. Use when starting a daemon on a Mac host, tuning its gates, or debugging boot refusals.

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No auth in v0: tailnet-only.

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Run the manzanasd daemon

One daemon per Mac, port :7433. Never run two daemons against the same CoreSimulator state; a second experimental daemon goes on 7434+ with --mock only.

Start

make build                                 # bin/manzanasd, bin/manzanas, bin/manzanas-broker
./bin/manzanasd --addr :7433                # on a Mac with Xcode
./bin/manzanasd --addr :7433 --mock         # anywhere (Linux/CI): mock fleet, no simctl
curl -s localhost:7433/v0/healthz          # {"ok":true,"version":"v0"}

Production installs use the LaunchAgent: deploy/install.sh --binary ../manzanasd (installs to ~/bin/manzanasd, RunAtLoad+KeepAlive, logs to ~/.manzanasd/logs/, log rotation). Upgrade = re-run install.sh with the new binary. Remove with deploy/uninstall.sh [--purge].

Flags that matter (each also has an MANZANASD_* env var)

FlagDefaultMeaning
--addr:7433listen address
--mockoffmock registry (no simctl; state/actions return 501)
--pool-simsemptycomma-separated UDIDs kept in the park/thaw warm pool
--pool-slim-profileemptysimslim profile applied to pool sims before parking (simslim v0.6.0+: batched launchctl transitions make the slim step much faster, esp. Intel; simslim verify gives exact drift checks; profiles needing in-app purchases keep store in their except array)
--pool-max-running0 = class default (2 Intel / 4 Apple Silicon)cap on Booted un-parked sims; negative disables
--pool-load-factor0 = 2.0refuse boots when 1-min load > factor x cores; negative disables
--pool-min-free-disk-gb0 = 5refuse boots below this free disk; negative disables
--pool-footprint-cap-mb0 = 3072watchdog recycles a pool sim above this phys_footprint; negative disables
--simbridge~/bin/simbridge or $PATHwarm-actions helper binary
--no-warmoffdisable the warm actions backend
--warm-max-targets / --warm-idle-ttl4 / 5mresident helper capacity / idle shutdown
--janitor-reap-stale-locksofflet the janitor shut down (never delete) unmanaged Booted sims whose sim-<udid> fleet lock is stale (>=2h) or missing, after the staleness persists a full grace (>=10 min); recommended ON for fleet daemons
--lock-dir/tmp/manzanas_locks (env MANZANAS_LOCK_DIR)fleet lock directory holding sim-<udid> lock files; fleet machines set MANZANAS_LOCK_DIR=/tmp/manzanas_locks
--journal-dir~/.manzanasd/journalempty string disables the journal
--journal-max-age / --journal-max-bytes168h / 2 GiBjournal GC bounds
--stream-max-streams / --stream-max-viewers / --stream-max-fps / --stream-lingersee --helpstreaming limits

Safety gates & 503 handling

On real (darwin, non-mock) daemons every boot passes: running-sim cap, load gate, free-disk gate. A refusal is 503 {"code":"overloaded"} — this is back-pressure, not an error: clients back off and retry (or lease on another host via the broker). Don't loosen the gates on shared boxes to make a task pass; parked pool sims don't count against the cap.

A boot may also surface 500 carrying the failure of a previous accepted boot that died in the background — just retry.

Warm pool

--pool-sims sims are booted, slimmed (--pool-slim-profile), then PARKED (SIGSTOPped, ~0 CPU). Lease grant thaws in ~25ms (AS) / ~225ms (Intel); release+reset erases, re-boots, re-parks. See the park-thaw-semantics skill before touching parked sims by hand.

Debugging

  • Logs: ~/.manzanasd/logs/manzanasd.{out,err}.log (LaunchAgent installs).
  • 501 not_implemented from state/actions endpoints = mock build or missing backend (e.g. no simctl on the host, journal disabled).
  • 503 unavailable from actions = AXe missing or the a11y bridge stayed unready — check ~/bin/axe exists and the sim is fully booted.
  • No auth in v0: tailnet-only. Never expose the port publicly.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.