SPB Git

Repository Public License

implementation-plan.md 4.6 KB 57 lines markdown

1<!--2ORIX OS Automation3File: docs/architecture/implementation-plan.md4Author: Simon-Pierre Boucher5Contact: contact@spboucher.ai6-->7# Implementation plan89Status of each milestone is tracked in `docs/implementation-status.md`; this file is the plan and does not change10when work lands.1112## Milestones and gates1314| # | Deliverable | Gate (how it is checked) |15|---|---|---|16| M0 | Audit, contracts (`ORIXCore`), credential broker, threat model, build/test scripts | `Scripts/build/test-all.sh` green for Core; security boundaries documented |17| M1 | App shell, design system, onboarding, settings, Keychain, SQLite store, console, permission center | App launches; settings persist across relaunch; Keychain integration tests pass |18| M2 | Three provider adapters, agent loop, budgets, cancel, approval UI | Fixture suites pass; one live smoke per provider; a real tool call through the executor |19| M3 | Filesystem, shell, git, http, documents, python, scripts, AppleScript/JXA, Accessibility, screen, browser | End-to-end tasks with runtime-checked postconditions |20| M4 | Workflow IR + codec + validator + engine, canvas, code editor, debugger, templates | Create → edit visually → edit as code → save → reopen → run, no semantic loss (round-trip tests) |21| M5 | Background agent, scheduler, file watchers, durable queue, export/import | A scheduled workflow runs with the window closed; missed-run policy tested with a manual clock |22| M6 | Delegation, scoped memory, skill packages, skill generation pipeline | An agent proposes a skill; it is analysed, tested confined, approved, installed, run |23| M7 | Watch & Learn recorder → proposed workflow, accessibility and performance passes | A recording becomes a replayable workflow after approval; secure fields never captured |24| M8 | Regression, security review, docs with real screenshots, signed + notarized DMG | `spctl` accepts the stapled artifact; evidence in `docs/release/` |2526## Test strategy2728- Each package: unit tests + integration tests against the real thing (files, subprocesses, SQLite, WKWebView,29  osascript). Suite names say `Fixture`/`Mock` when a double stands in for a provider.30- Live provider tests run only through the credential broker and print `SKIPPED(...)` without it.31- Tests needing a TCC permission check the real status and print what they skipped; they never prompt.32- Security tests map one-to-one to the ten threat scenarios (`docs/security/threat-model.md`).33- End-to-end: `orixctl` (command-line front end to the same services as the app) drives agent and workflow runs in a34  temp workspace; XCUITest drives the app itself.35- `Scripts/test/test-all.sh` runs every package and writes `docs/release/test-report.json` with counts per package36  and per class (unit / integration / live / skipped).3738## Risk register3940| Risk | Likelihood | Impact | Mitigation |41|---|---|---|---|42| A model is steered by content it reads (prompt injection) | High | High | Tool output framed as data; no capability comes from text; network/file tools gated by grants; consequential tiers need approval; tests per scenario |43| Secret reaches a prompt, log or artifact | Medium | High | Keys only in Keychain; resolved inside the executor; `Redactor` on events, results, receipts; release secret scan |44| `sandbox-exec` removed or changed by Apple | Low now | High | Profile verified by tests on this OS; script tools fail closed if confinement cannot be applied |45| TCC permission missing in the field | Certain | Medium | Non-prompting status checks, guided onboarding, `BLOCKED_PERMISSION` with the exact pane to open |46| Background agent acts with stale authority | Medium | High | Grants carry expiry/revocation; scheduled mode never auto-approves; revoke-all stops and clears |47| GUI automation hits the wrong window | Medium | High | Semantic AX targeting, re-resolve + focus check before every action, exclusive concurrency, emergency stop |48| Provider API drift (models, fields) | High | Medium | Model discovery at runtime; registry dated; adapters tolerate unknown fields; live smoke tests |49| Cost runaway | Medium | Medium | Per-run budgets enforced in the loop from reported usage; unknown prices flagged as incomplete, not guessed |50| Two processes corrupt shared state | Low | High | SQLite WAL + busy timeout, leases, per-run sequence assigned in a transaction, quarantine of a corrupt file |51| Scope: the brief is very large | Certain | Medium | Milestones in order, each usable; status file states what is unverified or missing rather than implying completeness |5253## Definition of done5455As in the brief (§23). Anything short of it is listed under "Not done" in `docs/implementation-status.md` with the56reason and what it needs.57
Files main
Menu

Theme