arm-syntheticslisted

# Brain Synthetics (observability surface 7) This surface runs **per-arm health monitoring**: each arm gets its own synthetics config, its own GH Action runner, and alerts that stay in the arm's own repo. Arm isolation is preserved — the brain never sees one arm's failures bleed into another arm's monitoring surface. ## When to use - Onboarding a new arm: drop the template files, edit `synthetics.yaml`, commit, scheduled probes start working - Existing arm needs health monitoring beyond what's already there (e.g. `sitemap-test` for the public site already exists for one arm — synthetics generalises it) - Operator says "monitor X for me daily" / "sondea Y cada hora" / "health check para Z" ## Files this skill installs Two templates live under `~/.claude/templates/arm-synthetics/`: 1. **`synthetics.yaml.template`** — copied into the arm at `<arm>/synthetics.yaml`. Operator edits the endpoints + expectations. 2. **`synthetics.yml.workflow-template`** — copied into the arm at `<arm>/.github/workflows/arm-synthetics.yml`. Reads the yaml, runs the probes, opens an issue ON THE ARM REPO on failure. A helper script `~/.claude/scripts/arm-synthetics-runner.py` does the actual probing — sits in the brain so all arms share the same runner code without duplication. ## Workflow when installing into a new arm 1. From inside `<arm>/` repo, run: ``` cp ~/.claude/templates/arm-synthetics/synthetics.yaml.template ./synthetics.yaml mkdir -p .github/workflows cp ~/.claude/te