Update to Autonom

{
  "description": "Reuse and record per-app Autonom knowledge for device work. Read the app's overlay and flow runbooks before driving a device, and save a new flow only when it is durable, reusable, and evidence-backed. Triggers on a device task against a known app (a package/bundle id is in play), and after working out a repeatable procedure — make an order, mock a screen, log in, reproduce a bug.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 332
    }
  ],
  "name": "mobile-memory",
  "skill_md_contents": "---\nname: mobile-memory\ndescription: Reuse and record per-app Autonom knowledge for device work. Read the app's overlay and flow runbooks before driving a device, and save a new flow only when it is durable, reusable, and evidence-backed. Triggers on a device task against a known app (a package/bundle id is in play), and after working out a repeatable procedure — make an order, mock a screen, log in, reproduce a bug.\n---\n\n# Mobile Memory\n\nPer-app knowledge that a device task needs but the CLI cannot know: the backend,\nhow the app's network is wired, where its response schemas live, and step-by-step\nflows worked out once and replayed after. It exists so the same hour of digging\nis not spent twice.\n\n## Where it lives\n\n```\n~/.autonom/apps/<package>/\n  app.md              identity, backend, network model, TLS, schema roots, gotchas\n  flows/<name>.md     a runbook: goal, endpoint, schema source, exact steps, verify\n  bodies/<name>.json  ready mock payloads that flows reference\n```\n\n**Machine-local and per-person**, plain markdown — not in the app repo, not in\nthe Autonom repo. The same idea as a personal memory store: it accumulates as you\nwork and stays yours. The skill (this file) ships with Autonom; the knowledge\ndoes not.\n\n## Before a device task\n\n1. Resolve the package id — from `--app-id`, the current session\n   (`autonom session show`), or the repository under test.\n2. Read `~/.autonom/apps/<package>/app.md` if it exists. It answers, up front,\n   the questions that otherwise surface as wasted time: which backend, does the\n   app cache the proxy, is `dart:io` even used, where are the schemas.\n3. If the task matches a flow (mock a screen, make an order), read\n   `flows/<name>.md` and follow it instead of re-deriving the endpoint, schema\n   and gotchas.\n4. Nothing there yet? Proceed normally — and consider leaving a flow behind when\n   you crack something that will recur.\n\n## When to save — and when not\n\nSave **only** knowledge that is durable, reusable, and evidence-backed: the exact\nendpoint, the exact glob, the path to the schema, the gotcha that cost time.\nPrefer updating the existing file over adding a near-duplicate.\n\nDo **not** save one-off exploration, obvious single commands, or anything already\nin the code. **Never** write secrets — reference \"log in as the test driver\", not\nan account or token. Keep each entry id-centric and concrete (`com.example.rides`,\n`POST /api/v1/orders`, `lib/models/**/*_response.dart`), never vague.\n\nA flow is worth a file when it is multi-step and will happen again. A single verb\nis not a flow.\n\n## Format\n\n- **app.md** — identity (package, backend, build), how the network works (native\n  vs `dart:io`, proxy caching), TLS/CA notes, where schemas live, a *known\n  screens* table (screen → endpoint → flow), and gotchas already handled by the\n  tool so a reader does not redo them.\n- **flows/<name>.md** — Goal · Endpoint · Schema source · Steps (exact CLI) ·\n  Verify · Gotchas. Reconstruct a response schema from the freezed model:\n  `@JsonKey(name: ...)` is the JSON field, `required` without `?` is mandatory.\n  These are prose runbooks — working knowledge. When one has become fully\n  deterministic (stable selectors, no judgment calls), propose converting it\n  into an Autonom Flow v1 file in the app repository (`mobile-flow`), which\n  the CLI can then validate and replay exactly.\n- **bodies/<name>.json** — a captured or hand-built payload a flow points at, so\n  the shape is not rebuilt each time.\n\n## How it connects\n\n`project-router` detects the app → this skill loads its overlay and any matching\nflow → the `mobile-session` / `mobile-screen` / `mobile-network` skills run the\nCLI verbs → new knowledge is written back here. The knowledge layer sits beside\nthe thin verb-skills; it does not replace them.\n\n## Related\n\n- `project-router` — routes here once a device task targets a known app.\n- `mobile-session`, `mobile-screen`, `mobile-network` — the verbs a flow drives.\n"
}