backlog.json

{
  "schema_version": 1,
  "field_semantics": {
    "status": "initial_historical_planning_snapshot",
    "agent": "initial_historical_planning_snapshot",
    "live_dispatch_authority": "GitHub Project #2 and current issue bodies",
    "do_not_redispatch_from_snapshot": true
  },
  "issues": [
    {
      "key": "G01",
      "title": "Prove Scale Set delivery, acquisition and drain contracts",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [],
      "goal": "Select and pin a supported Scale Set integration path with demonstrated recovery at message acknowledgement, acquisition and JIT boundaries.",
      "scope": "Isolated protocol spike and ADR; compare release v0.4.0 with audited commit cb0405b2d874500e75ae34eff8d582ab75956b45. Do not implement production providers or copy the demo shutdown policy.",
      "tdd": [
        "Red: fake session server reproduces ACK-before-callback crash and missing lifecycle callback.",
        "Cover pre/post ACK, acquisition, >50 events, repeated/stale statistics, 202, 401/403/429, callback failure and session replacement with deterministic barriers.",
        "Test idle-assignment races while drain withdraws capacity; mocks alone cannot prove server deregistration/acquisition behavior. Run an explicitly authorized disposable private live canary before closing the gate."
      ],
      "acceptance": [
        "Record exact module/runner versions and supported JIT bootstrap transports, including any secret exposure.",
        "Choose high-level listener plus independent reconciliation OR a lower-level supported message loop; do not simply reverse ordering without evidence.",
        "Document recoverable state versus quarantine for lost callback/request IDs; no exactly-once claim.",
        "Publish sanitized live contract evidence or keep the gate unresolved; obtain independent Luna max protocol review."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Ready",
      "test_profile": [
        "offline",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 1,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/1"
      },
      "release": "R1 - Internal MVP"
    },
    {
      "key": "G02",
      "title": "Prove local App enrollment and launchd credential access",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [],
      "goal": "Resolve GitHub App Manifest loopback enrollment and the macOS credential/service identity contract with real platform evidence and a safe manual-import fallback.",
      "scope": "Gate experiments and ADR only. Use disposable authorized Apps/org installations; do not alter production App settings, existing runners or disable FileVault.",
      "tdd": [
        "Red: reject wrong/expired/replayed state, duplicate callback parameters, wrong Host and forged installation IDs in a small harness.",
        "Live-check disabled-webhook Manifest shape, HTTP loopback random-port redirect and two org installations; distinguish OAuth from Manifest.",
        "Test Keychain read under actual signed/source-build launchd identity, screen lock, login/logout and reboot; identify file-based Keychain requirement for a system daemon."
      ],
      "acceptance": [
        "Confirm minimal org runner permissions, organization-owner installation requirement and per-org identity checks.",
        "Decide login-scoped startup and dedicated controller/job identities; document cold-boot limitations and any narrow helper.",
        "Manual App import works if Manifest is rejected; partial setup never silently creates extra Apps.",
        "Publish sanitized results and explicit support matrix; no secret values or raw callback URLs/codes."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Ready",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 2,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/2"
      },
      "release": "R3 - General distribution",
      "r1_subset": "G02-R1"
    },
    {
      "key": "G03",
      "title": "Bootstrap Go module and independent public CI",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [],
      "goal": "Establish a reproducible Go development and TDD baseline whose public PR checks run without gh-runnerd, live GitHub credentials or local runner access.",
      "scope": "Supported Go toolchain pin, module/package skeleton, formatting/vet/unit/race/fuzz-smoke/vulnerability checks and contribution commands; no runtime behavior.",
      "tdd": [
        "Verify a negative CI smoke in a temporary fixture (for example malformed Go fails formatting/compilation); do not invent application behavior for scaffolding. Meaningful application red/green evidence begins with G04 contracts.",
        "Verify ordinary checks succeed with no App credentials and no Docker daemon.",
        "Review workflow permissions and trusted/untrusted triggers; pin external actions to reviewed commits."
      ],
      "acceptance": [
        "Record toolchain and dependency policy, licence inventory and local commands.",
        "CI uses hosted standard environments and cannot bootstrap-deadlock on this manager.",
        "Skipped hardware/live suites are explicit and never represented as passed.",
        "No product daemon is installed by tests."
      ],
      "risk": "Medium",
      "kind": "implementation",
      "priority": "P1",
      "status": "Ready",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 3,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/3"
      },
      "release": "R1 - Internal MVP"
    },
    {
      "key": "G04",
      "title": "Freeze CLI configuration, provider and scaling contracts",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [
        "G01",
        "G02",
        "G03"
      ],
      "goal": "Publish unambiguous versioned contracts for CLI configuration, provider capabilities, resource ownership and fixed/autoscale policies before parallel implementation.",
      "scope": "Design/schema fixtures and interface contracts. Define fixed count and autoscale minTotal/maxTotal; defer idleBuffer. Include explicit existing Docker endpoint and trust domain.",
      "tdd": [
        "Red: invalid/negative limits, min greater than max, incompatible provider capabilities and duplicate pool IDs are rejected by contract fixtures.",
        "Specify desired=min(maxTotal,max(minTotal,TotalAssignedJobs)); reducing below busy count affects admission only.",
        "Table-test serialized request/response versions, unknown state and resource identity including process birth evidence."
      ],
      "acceptance": [
        "CLI --min maps to minTotal, not minimum idle workers; all docs/examples agree.",
        "Define provider create/discover/inspect/retire/logs capabilities and owned-resource identity.",
        "Document native controller/job identity and Docker DinD privilege profiles; an adapter cannot silently broaden trust.",
        "Assign shared file ownership; independent provider implementation can proceed without redefining contracts.",
        "Define runner freshness/security update ownership, staged refresh and rollback; a prohibited stale version must not receive new jobs."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 4,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/4"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G05",
      "title": "Implement durable state and operation journal",
      "stage": "M1 - Control plane",
      "agent": "Luna max",
      "dependencies": [
        "G03",
        "G04"
      ],
      "goal": "Persist desired/observed worker state and side-effect intents so interrupted operations reconcile without uncontrolled duplicate workers or foreign-resource deletion.",
      "scope": "SQLite schema/migrations, typed reducer and ownership ledger; no live provider side effects.",
      "tdd": [
        "Red: reopen a real temporary DB after kill points at intent commit, reservation, lost provider response and cleanup-before-commit.",
        "Property/fuzz tests reject invalid transitions and keep unknown workers reserved.",
        "Inject DB busy, disk-full, interrupted migration and PID/resource reuse; inspect old/new state after reopen."
      ],
      "acceptance": [
        "Stable identities include controller/org/pool/generation/immutable provider ID.",
        "Intent is durable before side effect; retries are idempotent and ambiguous outcomes quarantine.",
        "No credentials/JIT payload in DB; migrations have tested rollback/recovery policy.",
        "Crash tests do not claim to simulate physical power loss unless actually tested."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 5,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/5"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G06",
      "title": "Implement the supervised daemon and authorized local IPC",
      "stage": "M1 - Control plane",
      "agent": "Luna max",
      "dependencies": [
        "G05"
      ],
      "goal": "Run one foreground daemon under supervision and authorize CLI commands through a private versioned Unix socket with durable operation identities.",
      "scope": "Lock ownership, private socket, peer UID, server identity, bounded message parsing and durable mutations; no remote network control API.",
      "tdd": [
        "Red: duplicate daemon, wrong peer UID, job UID invoking the real CLI, wrong socket owner and symlink substitution cannot mutate state.",
        "Test concurrent/replayed commands, stale epochs, invalid methods and oversized frames.",
        "Kill daemon while clients are connected; reconnection returns accurate committed/uncommitted outcomes."
      ],
      "acceptance": [
        "Terminal/CLI exit does not stop managed workers.",
        "No arbitrary shell/path/mint-token endpoint exists; every mutation is authorized.",
        "A daemon restart cannot create a second owner or adopt manual runners.",
        "No job-user access to control socket in the supported profile."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime"
      ],
      "github_issue": {
        "number": 6,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/6"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G07",
      "title": "Implement GitHub App credential and installation binding",
      "stage": "M1 - Control plane",
      "agent": "Luna max",
      "dependencies": [
        "G02",
        "G05"
      ],
      "goal": "Authenticate each configured organization with a verified user-owned App installation while keeping management secrets local and out of all worker/control outputs.",
      "scope": "Credential-store adapter, App import, canonical org/App binding, SDK refresh facade and safe rotation; no auto browser creation.",
      "tdd": [
        "Red: wrong org/App/personal/suspended/underprivileged installation is refused; only configured installations are used.",
        "Fake full SDK auth chain, concurrent expiry, clock skew, 401 refresh, 403 revoke and per-org isolation.",
        "Seed secret canaries into raw SDK response bodies/URLs/errors; verify absence from logs/status/DB/backend spawn.",
        "Locked/denied Keychain fails closed; crash during key/profile swap preserves a recoverable old/new selection."
      ],
      "acceptance": [
        "Baseline org permission is organization_self_hosted_runners:write; no repo-admin escalation.",
        "One org auth failure stops its new acquisition but does not disable the other healthy org.",
        "Shared App compromise blast radius and key-memory limitations are documented.",
        "Actual G02 credential identity contract is implemented; no plaintext fallback."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 7,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/7"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G08",
      "title": "Implement guided init and optional Manifest enrollment",
      "stage": "M1 - Control plane",
      "agent": "Luna max",
      "dependencies": [
        "G06",
        "G07"
      ],
      "goal": "Provide a resumable init flow that imports or creates an operator-owned App and binds authorized organizations without copying per-runner registration tokens.",
      "scope": "CLI onboarding over the G02-reviewed auth contract. Manual import is first-class; Manifest available only if its live gate passed.",
      "tdd": [
        "Red: denial, cancelled browser, port conflict, timed-out conversion, replay and a crash between remote creation/key commit leave honest partial status.",
        "Never convert a callback more than once or trust browser installation_id without API verification.",
        "Golden CLI tests use synthetic data; no PEM/code/key material in output or failure diagnostics."
      ],
      "acceptance": [
        "Init does not require an external callback server or gh OAuth session.",
        "Pending second-org installation can be resumed without duplicating the first.",
        "App permissions/trust/startup expectations are shown before user browser approval.",
        "Luna max reviews all auth-facing code and evidence before merge."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 8,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/8"
      },
      "release": "R3 - General distribution"
    },
    {
      "key": "G09",
      "title": "Implement the recoverable Scale Set adapter",
      "stage": "M1 - Control plane",
      "agent": "Luna max",
      "dependencies": [
        "G01",
        "G05",
        "G07"
      ],
      "goal": "Translate verified Scale Set demand and lifecycle observations into durable local intents with bounded retry and explicit advertised capacity.",
      "scope": "One client/session per pool/org, JIT attempt binding, safe errors, reconnect and reviewed ACK strategy; no provider-specific lifecycle policy.",
      "tdd": [
        "Red: crash after ACK/before callback and lost JIT response cannot permanently lose recoverable capacity or reuse abandoned JIT.",
        "Replay >50 jobs, stale messages, expired sessions, 429 backoff and prolonged outage with a fake clock.",
        "Confirm aggregate stats drive demand while unknown job identity remains conservative."
      ],
      "acceptance": [
        "Uses G01 pinned integration and real evidence, not the instructional Docker example lifecycle.",
        "No raw SDK error logs, cross-org mutable client state or unbounded retries.",
        "Stops new acquisition on authorization loss and reconciles before resuming.",
        "Owned scale-set/runner deletion requires durable identity, not a name prefix."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 9,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/9"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G10",
      "title": "Implement disposable Linux workers with isolated Docker services",
      "stage": "M2 - Execution and scaling",
      "agent": "Luna max",
      "dependencies": [
        "G04",
        "G05",
        "G07"
      ],
      "goal": "Provision ARM64 one-job Linux workers against an explicit existing Docker engine with worker-owned Docker services, resources and cleanup.",
      "scope": "Docker adapter, pinned images, per-worker daemon/network/storage and logs. No automatic Docker Desktop/Lima installation or global context mutation.",
      "tdd": [
        "Red: common provider contract for create/lost response/retry/inspect/drain/cleanup preserves unrelated containers and volumes.",
        "Run ordinary shell, checkout-like mounts, job container and simultaneous PostgreSQL/Redis service fixtures on real ARM64 Docker.",
        "Probe job mounts/env/engine inspection for manager secrets and management socket exposure; enforce the reviewed trusted DinD profile.",
        "Runtime outage and partial cleanup retain reservations/quarantine instead of broad prune."
      ],
      "acceptance": [
        "DB ports/mount paths and resource ceilings include service daemon overhead.",
        "Images pinned by digest and ARM64 manifest, required runtime tool dependencies documented.",
        "Dedicated job Docker daemon only; privileged DinD if needed is an explicit trusted-runtime exception, not hostile-code isolation.",
        "Existing manual runners/runtime objects remain untouched.",
        "Refresh only future workers to a verified supported runner/image version; preserve busy workers and block stale admission."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 10,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/10"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G11",
      "title": "Implement trusted native macOS worker identity and lifecycle",
      "stage": "M2 - Execution and scaling",
      "agent": "Luna max",
      "dependencies": [
        "G02",
        "G04",
        "G05",
        "G07"
      ],
      "goal": "Run one-job macOS ARM64 workers under the reviewed job identity with correct process ownership, recoverable supervision and scoped cleanup.",
      "scope": "Dedicated controller/job identity integration and minimal allowlisted helper if required by G02, process-group lifecycle, unique workdir and logs. No VM isolation claim.",
      "tdd": [
        "Red: job identity cannot read controller key/state/socket; same-UID mode is unsupported by the protected default.",
        "A long-running worker survives manager restart; PID reuse cannot authorize killing another process.",
        "Inject cancellation during bootstrap, surviving child processes, workdir symlink, cleanup failure and conflicting host-tool mutation.",
        "JIT transport follows verified runner support, with residual in-job exposure documented and no management key inheritance."
      ],
      "acceptance": [
        "Native provider is enabled only for explicit trusted repositories/trust domain.",
        "Different organizations do not share a native job identity or writable credential/cache areas by default.",
        "Helper accepts only bounded approved operations, not arbitrary executable paths, UIDs or shell text.",
        "Unsupported identity/Keychain/service conditions fail preflight; fresh workdir is not advertised as fresh OS.",
        "Controller owns pinned runner download/update policy; verify checksum and refresh new workers without replacing the binary beneath busy workers."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 11,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/11"
      },
      "release": "R3 - General distribution"
    },
    {
      "key": "G12",
      "title": "Implement deterministic shared capacity and fair scheduling",
      "stage": "M2 - Execution and scaling",
      "agent": "Luna max",
      "dependencies": [
        "G04",
        "G05"
      ],
      "goal": "Compute deterministic fixed/autoscale targets and serial admission proposals that respect per-pool, per-runtime and host-wide budgets, with fair progress across eligible pools.",
      "scope": "Pure policy module with fake clock and immutable typed snapshots. Luna max-owned G04 defines policy; G05 owns atomic reservation persistence and G13 owns live serialized application. No concurrent side effects, protocol acquisition or provider mutations in this issue.",
      "tdd": [
        "Red: a pure reduction of simultaneous two-org demands proposes at most one allocation for the last CPU/memory slot; stale or duplicate proposal application is exercised by G05/G13 integration, not claimed solved here.",
        "Property tests cover zero/max/minTotal, infeasible floors, busy-over-reduced-cap, creation/cleanup/unknown reservations and pressure cutoff.",
        "Finite scenario tests prove no starvation as slots free, deterministic tie-breaking and no oscillation under chosen backoff."
      ],
      "acceptance": [
        "Uses fixed count or minTotal/maxTotal semantics consistently; explicit idleBuffer deferred.",
        "Native memory budget is admission-only; no hard OS cap promise.",
        "Existing engine allocation and daemon/OS headroom are represented without double-counting.",
        "Luna max reviews invariants/property counterexamples before connecting live effects.",
        "Return versioned proposals, never mutate reservations or spawn workers; G13 Luna max integration rejects stale proposals and commits capacity atomically."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 12,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/12"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G12a",
      "parent": "G12",
      "title": "Implement pure scaling targets and capacity arithmetic",
      "stage": "M2 - Execution and scaling",
      "agent": "Luna max",
      "dependencies": [
        "G03"
      ],
      "goal": "Implement deterministic, overflow-safe arithmetic for desired runner totals and additional resource capacity as an independent leaf of G12.",
      "scope": "Three stdlib-only pure calculations under the independently reviewed numeric contract: fixed target, autoscale target and additional capacity. No config/provider schema, durable reservation, proposal application, fairness, I/O or live effects. G04/G12 integration gates remain required.",
      "tdd": [
        "Capture a behavioral red for this new feature, not a compilation-only result or a claimed existing product bug.",
        "Cover zero/negative/min-max limits, exact last slot, reduced limits below committed usage, distinct CPU/memory and host/runtime constraints, and invalid later input after an early zero result.",
        "Cover permitted individual zero-cost host budget plus positive runtime cost, reject nonempty all-zero cost set even for zero request, and independently check MaxInt64 behavior with math/big feasibility/maximality and bounded fuzz."
      ],
      "acceptance": [
        "Three pure calculations implement the approved numeric contract without overflow or input mutation.",
        "No new dependency or live integration; future caller accounting, stale calculation and atomic application limits are explicit.",
        "Independent Astra contract/counterexample review and exact-head GitHub Codex review resolved before merge.",
        "Completing this child does not complete G12 or waive G01/G02/G04/G05 integration gates."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P1",
      "status": "Ready",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 40,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/40"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G13",
      "title": "Integrate reconciliation, drain and safe restart",
      "stage": "M2 - Execution and scaling",
      "agent": "Luna max",
      "dependencies": [
        "G06",
        "G09",
        "G10",
        "G11",
        "G12"
      ],
      "goal": "Reconcile GitHub and provider observations into safe worker lifecycle changes across crashes, scale-down and network/runtime failures.",
      "scope": "Connect durable intents to providers and scheduler, stop/resume/restart/drain and quarantine; no automatic job rerun or force-delete default.",
      "tdd": [
        "Red: assignment immediately after idle observation cannot be killed during scale-down.",
        "Inject every intent/side-effect/result-commit crash window, message loss, one-org failure and Docker disconnect.",
        "Test drain timeout reports busy/unknown workers without force kill; repeated commands survive restart.",
        "Host reboot reports interrupted jobs accurately; manual/foreign resources never adopted."
      ],
      "acceptance": [
        "Reservations released only after known terminal/cleanup state.",
        "Manager restart preserves surviving workers and quarantines ambiguous state.",
        "Future capacity withdrawn during drain without claiming atomic remote cancellation.",
        "Luna max independently reviews end-to-end failure matrix."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime"
      ],
      "github_issue": {
        "number": 13,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/13"
      },
      "release": "R2 - Everyday operations",
      "r1_subset": "G13-R1"
    },
    {
      "key": "G14",
      "title": "Implement status, logs and sanitized diagnostics",
      "stage": "M3 - Reliability qualification",
      "agent": "Luna max",
      "dependencies": [
        "G06",
        "G13"
      ],
      "goal": "Give operators one accurate view of organizations, pools, capacity, jobs and failures with bounded redacted local logs and machine-readable output.",
      "scope": "status/list/describe/logs/doctor/export and telemetry adapters; no new auth scopes unless justified by an approved feature.",
      "tdd": [
        "Red: desired/reserved/creating/busy/draining/unknown counts reflect durable and stale observations separately.",
        "Test concurrent tail/rotation, slow clients, missing provider and bounded diagnostic export.",
        "Secret canaries in SDK errors, config and job data never appear in CLI/JSON/logs/metrics."
      ],
      "acceptance": [
        "Logs survive worker cleanup with retention/disk limits.",
        "Metrics distinguish manager overhead, engine limits and job costs.",
        "Stale/offline state is visible rather than falsely healthy.",
        "Exit codes/JSON version and cancellation behavior are documented."
      ],
      "risk": "Medium",
      "kind": "implementation",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime"
      ],
      "github_issue": {
        "number": 14,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/14"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G15",
      "title": "Implement startup, shutdown and safe configuration updates",
      "stage": "M3 - Reliability qualification",
      "agent": "Luna max",
      "dependencies": [
        "G02",
        "G06",
        "G13"
      ],
      "goal": "Install and manage the daemon through launchd with honest login/boot behavior, safe config generations and reversible lifecycle commands.",
      "scope": "startup/unstartup/restart/doctor installation and config persistence. Do not promise cold-boot access unless G02 established it.",
      "tdd": [
        "Red: CLI without an interactive shell/ambient env uses correct endpoint/keychain paths.",
        "Test stale lock, duplicate launch registration, malformed config, generation update mid-job, rollback and daemon crash restart.",
        "Run actual supported login/lock/reboot mode; uninstall does not automatically delete Apps or legacy workers."
      ],
      "acceptance": [
        "launchd supervises foreground daemon; terminal closure irrelevant.",
        "Busy jobs survive ordinary manager upgrades; unsupported service version rejected safely.",
        "Explicit boundaries among daemon uninstall, owned-resource cleanup, local key deletion and remote App uninstall.",
        "Luna max reviews credential/service identity and cleanup transitions."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime"
      ],
      "github_issue": {
        "number": 15,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/15"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G16",
      "title": "Build repeatable two-organization runtime qualification",
      "stage": "M3 - Reliability qualification",
      "agent": "Luna max",
      "dependencies": [
        "G08",
        "G13",
        "G14",
        "G15"
      ],
      "goal": "Provide a repeatable, opt-in hardware/live test suite proving real GitHub workflow routing, Linux DB services and macOS jobs across two organizational installations.",
      "scope": "Fixtures, scripts and sanitized evidence format from reviewed contracts; no unauthenticated public PR access to local hardware.",
      "tdd": [
        "Red: tests detect wrong org/label routing, leftover DB/container/workdir and busy-job termination.",
        "Execute Linux shell/job-container/services workflows and native macOS workflows, queue bursts and cap changes.",
        "Use immutable reviewed commit and disposable controlled resources; record every skipped profile."
      ],
      "acceptance": [
        "Both orgs and both platforms appear in evidence, with actual job URLs sanitized if private.",
        "Parallel service tests catch port/network/workspace conflicts.",
        "Cleanup uses owned inventory and preserves existing manual infrastructure.",
        "Luna max signs off contract coverage; credentials stay outside fixtures/artifacts."
      ],
      "risk": "Medium",
      "kind": "implementation",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 16,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/16"
      },
      "release": "R3 - General distribution",
      "r1_subset": "G16-R1"
    },
    {
      "key": "G17",
      "title": "Qualify crash recovery, soak and resource budgets",
      "stage": "M3 - Reliability qualification",
      "agent": "Luna max",
      "dependencies": [
        "G16",
        "G19"
      ],
      "goal": "Establish release-candidate reliability and measured resource limits through deterministic faults and a real mixed-workload soak.",
      "scope": "24-hour/100-job proposed release gate, fault matrix, performance baseline and decision report. Do not promote if required evidence is missing.",
      "tdd": [
        "Kill daemon at all identified barriers, interrupt network/runtime, rotate tokens and simulate disk pressure.",
        "Run queue bursts and repeated drain/scale transitions with both orgs and OS pools.",
        "Measure manager CPU/RSS, API traffic, admission/ready latency, queue time and disk growth on fixed hardware.",
        "Control host reboot and classify expected interrupted jobs separately from manager recovery faults.",
        "Simulate stale runner/image and urgent required update during busy work; prove paused new admission, safe refresh/drain and no rollback to forbidden stale version."
      ],
      "acceptance": [
        "No unexplained duplicates/orphans/cap violations/busy kills/secret exposure.",
        "Set recommended counts from measured headroom, not host marketing specs.",
        "Document SDK preview limits and residual trust/platform limitations.",
        "Independent Luna max reliability verdict; incomplete soak remains incomplete."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 17,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/17"
      },
      "release": "R3 - General distribution"
    },
    {
      "key": "G18",
      "title": "Package signed releases and reproducible distribution",
      "stage": "M4 - Pilot and release",
      "agent": "Luna max",
      "dependencies": [
        "G03",
        "G15",
        "G17"
      ],
      "goal": "Produce verifiable macOS ARM64 release artifacts with install/upgrade/rollback instructions and accurate preview/stability labeling.",
      "scope": "Build packaging/checksums/provenance/SBOM/license notices/Homebrew formula plan; signing/notarization only with user-owned credentials when available.",
      "tdd": [
        "Red: corrupted artifact/signature and incompatible state version fail before replacement.",
        "Test clean install and upgrade rollback without deleting user state/keys or interrupting busy jobs.",
        "Verify PR builds require no signing secrets; release evidence records unsigned/notarized status truthfully."
      ],
      "acceptance": [
        "No downloaded macOS image, restricted virtualization binary or embedded App private key.",
        "No auto-publish/signing purchase implied; repository owner controls release channel.",
        "Bootstrap build remains independent of gh-runnerd.",
        "Luna max reviews supply-chain, update and rollback effects."
      ],
      "risk": "Medium",
      "kind": "implementation",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime"
      ],
      "github_issue": {
        "number": 18,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/18"
      },
      "release": "R3 - General distribution"
    },
    {
      "key": "G19",
      "title": "Audit trust admission and cross-worker secret boundaries",
      "stage": "M3 - Reliability qualification",
      "agent": "Luna max",
      "dependencies": [
        "G08",
        "G10",
        "G11",
        "G13"
      ],
      "goal": "Verify and enforce the advertised trust profiles so untrusted work cannot enter trusted-only pools and jobs cannot reach manager credentials/control in the supported configuration.",
      "scope": "Threat-model closure, policy preflight, native UID and Docker DinD profiles, adversarial non-destructive probes. No unsupported claim of hostile-container isolation.",
      "tdd": [
        "Red: mismatched runner-group access, ambiguous fork provenance and job invoking official CLI cannot bypass policy.",
        "Try reading controller home/state/keychain/socket, sibling files/env and manager Docker socket from controlled test jobs.",
        "Exercise SDK secret-in-error, proxy inheritance, wrong-org endpoint, unsafe mount/device/host-network requests.",
        "Test shared-App blast radius/rotation docs and reject unsupported arbitrary backend flags."
      ],
      "acceptance": [
        "Runner-group restrictions verified before new admission; no automatic broadening to all repos.",
        "Outer runner launches unprivileged; only reviewed worker-specific DinD has manager-runtime privilege. Trusted jobs control inner daemon; no claim that inner privileged containers are prohibited.",
        "Same-user native insecure mode is not protected default; unresolved boundaries block release.",
        "Public-fork support deferred until isolated VM/network provenance design is proven."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 19,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/19"
      },
      "release": "R3 - General distribution"
    },
    {
      "key": "G20",
      "title": "Pilot migration with reversible legacy runner handoff",
      "stage": "M4 - Pilot and release",
      "agent": "Luna max",
      "dependencies": [
        "G17",
        "G18"
      ],
      "goal": "Demonstrate a reversible rollout from existing manual runners to gh-runnerd-managed pools across both organizations without accidental interruption or resource adoption.",
      "scope": "Runbook and operator-authorized canary rollout. Existing manual services remain until canary success; no unattended production rewrite.",
      "tdd": [
        "Red: migration checks detect workflow labels still targeting the wrong pool or a missing access grant.",
        "Move one reviewed test workflow first, then expand; verify rollback can route back to old runners.",
        "Drain old owned-by-operator services only in the explicit handoff procedure after busy work completes."
      ],
      "acceptance": [
        "Linux and macOS canary evidence and pool resource settings are recorded.",
        "Legacy identities are never imported solely by name prefix.",
        "Workflow routing labels remain stable when scaling worker count.",
        "Record rollout verdict and rollback path; current planning task does not perform migration."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P1",
      "status": "Backlog",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 20,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/20"
      },
      "release": "R2 - Everyday operations"
    },
    {
      "key": "G21",
      "title": "Evaluate optional macOS VM and multi-host providers",
      "stage": "M5 - Future",
      "agent": "Luna max",
      "dependencies": [
        "G17"
      ],
      "goal": "Decide whether optional isolated macOS VM and multi-host adapters can meet safety, licensing and operating-cost requirements without coupling the core to one runtime.",
      "scope": "Research/prototype ADR only after core qualification. Evaluate Apple Virtualization helper and optional external tools; no default Tart bundling.",
      "tdd": [
        "Reproduce image clone/boot/one-job/teardown and controller-restart behavior in an isolated experiment.",
        "Measure startup/capacity and validate exact macOS VM licence/runtime limits separately from runner-process counts.",
        "Probe persistence, credential access and host/LAN networking before claiming untrusted-job support."
      ],
      "acceptance": [
        "Document go/no-go, license boundaries and optional provider interface changes.",
        "No unlimited scaling claim on one Mac or inference that all Linux VMs share macOS guest cap.",
        "Remote-host design includes trust/auth/recovery and central budget semantics before implementation.",
        "Scope stays outside v0.1; create bounded follow-up implementation issues only after evidence."
      ],
      "risk": "High",
      "kind": "research",
      "priority": "P2",
      "status": "Future",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 21,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/21"
      },
      "release": "Future research"
    },
    {
      "key": "G01f",
      "parent": "G01",
      "title": "Finalize paired execution with bounded owned cleanup",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [],
      "release": "R1 - Internal MVP",
      "goal": "Complete one private, same-invocation normal-success terminal path after issue #52's paired execution. Prove exact session closure, non-force worker deletion with its separate absence receipt, and owned scale-set deletion through real C/W journals and private SDK/TLS and Docker/Unix fixtures. This child does not authorize a live run, release admission claims or complete parent #1.",
      "scope": "Tracked on issue #54 (Done). This keyed record exists so G01g's native blockedBy #54 is resolvable in backlog.json. Do not duplicate runtime evidence here.",
      "tdd": [
        "See issue #54. Do not edit G01f evidence from the #66 documentation branch."
      ],
      "acceptance": [
        "Issue #54 remains the source of acceptance and is Done.",
        "Completing G01f does not complete G01 #1."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Done",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 54,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/54"
      }
    },
    {
      "key": "G01g",
      "title": "Wire paired terminal execution through a bounded broker handoff",
      "stage": "M0 - Evidence gates",
      "agent": "Luna max",
      "dependencies": [
        "G01f"
      ],
      "release": "R1 - Internal MVP",
      "goal": "Connect the reviewed paired terminal path to one same-process experimental executable and a narrowly bound broker handoff, with offline evidence that authority, credential separation and fail-stop recovery boundaries are preserved.",
      "scope": "Tracked on issue #60 and PR #62 in a separate worktree. This keyed record exists so R1 dependencies resolve. It is not G01 completion and does not duplicate runtime or CI evidence.",
      "tdd": [
        "See issue #60. Do not edit #60/#62 from the #66 documentation branch."
      ],
      "acceptance": [
        "Issue #60 remains the source of acceptance.",
        "Completing G01g does not complete G01 #1."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "In progress",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 60,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/60"
      }
    },
    {
      "key": "P66",
      "title": "Reframe delivery into MVP, operations and distribution releases",
      "stage": "M0 - Evidence gates",
      "agent": "Grok 4.6 xhigh",
      "review_agent": "Luna max",
      "dependencies": [],
      "release": "R1 - Internal MVP",
      "goal": "Reorganize delivery into an internal single-host MVP, an everyday operations release, and a general-distribution release without weakening the existing Scale Set recovery objective or safety/review gates.",
      "scope": "Documentation and mapping only. Preserve G01 issue #1 exact Goal and acceptance evidence. Do not create another native goal. No live operations. No production implementation of dependency-blocked issues.",
      "tdd": [
        "Prose-only planning change: use link checks, diff review and issue/Project mapping verification. Do not invent application tests."
      ],
      "acceptance": [
        "Three release stages and concrete user-visible exit criteria documented.",
        "All current issues mapped, with native dependency updates only where justified and no false Done states.",
        "Minimum executable MVP critical path identified, with unimplemented CLI/daemon/install work explicit.",
        "Existing G01 full objective preserved; no live evidence represented as complete.",
        "Focused independently reviewed PR, exact-head clean Codex verdict and CI before merge."
      ],
      "risk": "Medium",
      "kind": "planning",
      "priority": "P0",
      "status": "In progress",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 66,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/66"
      }
    },
    {
      "key": "G02-R1",
      "parent": "G02",
      "title": "Prove manual single-organization credentials for the foreground MVP",
      "stage": "M0 - Evidence gates",
      "agent": "Grok 4.6 xhigh",
      "review_agent": "Luna max",
      "dependencies": [],
      "release": "R1 - Internal MVP",
      "goal": "Demonstrate a manually configured single-organization credential path for the foreground MVP that binds the intended installation and never exposes management credentials to workers.",
      "scope": "Extract the R1 credential subset of #2 without closing or rewriting its broader Manifest, multi-organization or launchd qualification. Foreground-only; unattended login/logout/reboot service behavior is not supported in R1. No implicit live authorization.",
      "tdd": [
        "Red: missing/wrong/expired credentials, wrong installation and partial setup refuse safely without extra Apps or leaked values.",
        "Management credentials never reach worker argv/environment/files; only per-worker JIT follows the reviewed G01 transport.",
        "Live/runtime checks require reviewed commits and explicit concrete maintainer authorization."
      ],
      "acceptance": [
        "Explicit minimal permission, org/repository/installation identity and credential source validation before remote effects.",
        "Actual authorized manual configuration works; negative cases refuse safely.",
        "Credentials are not persisted or imported to Keychain implicitly.",
        "Broader #2 remains open until its own criteria pass."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Ready",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 67,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/67"
      }
    },
    {
      "key": "G13-R1",
      "parent": "G13",
      "title": "Deliver one foreground command for a bounded single-worker lifecycle",
      "stage": "M2 - Execution and scaling",
      "agent": "Grok 4.6 xhigh",
      "review_agent": "Luna max",
      "dependencies": [
        "G01g",
        "P66",
        "G02-R1"
      ],
      "release": "R1 - Internal MVP",
      "goal": "Provide a foreground MVP command that connects the reviewed broker and paired worker lifecycle for one organization, one private repository and one worker, with durable fail-stop state and safe owned cleanup.",
      "scope": "Vertical R1 slice of #4/#5/#6/#7/#9/#10/#13/#14 with primary parent #13. Reuse the reviewed paired Linux-container execution path. No second backend, native macOS jobs, daemon/install or context switch. Native dependencies remain G01g/#60, P66/#66 and G02-R1/#67. Until full G01 #1 and G02 #2 evidence gates pass, authorized work is a reviewed minimal contract and offline evidence only; production implementation starts only then. Completing G02-R1/#67 does not complete G02. Full G02 remains classified R3; user-visible R1 scope is not independently deliverable until that pre-release gate passes.",
      "tdd": [
        "Native blockers G01g/#60, P66/#66 and G02-R1/#67 remain the GitHub blockedBy set; do not add #1/#2 as native edges from this docs change.",
        "Until full G01 #1 and G02 #2 pass, work is contract/offline evidence only; production implementation is unauthorized.",
        "Negative tests for authority mutation, partial state, timeout, cancellation and non-owned cleanup.",
        "No claim that the existing empty product entry point already implements the command."
      ],
      "acceptance": [
        "One documented user-facing entry point, explicit configuration and safe status output.",
        "Bound org/repository/installation and one owned worker, shared capacity hard-limit one.",
        "Foreground interruption stops new admission and preserves busy/unknown work.",
        "Completing this child does not complete #4\u2013#15.",
        "Completing G02-R1/#67 does not complete G02. Production implementation waits for full G01 #1 and G02 #2."
      ],
      "risk": "High",
      "kind": "implementation",
      "priority": "P0",
      "status": "Blocked",
      "test_profile": [
        "offline"
      ],
      "github_issue": {
        "number": 68,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/68"
      }
    },
    {
      "key": "G16-R1",
      "parent": "G16",
      "title": "Qualify the internal MVP with an authorized real job and recovery evidence",
      "stage": "M3 - Reliability qualification",
      "agent": "Grok 4.6 xhigh",
      "review_agent": "Luna max",
      "dependencies": [
        "G01",
        "G13-R1"
      ],
      "release": "R1 - Internal MVP",
      "goal": "Qualify the single-host internal MVP with an explicitly authorized real private-repository job, safe cleanup and demonstrated ACK, acquisition and JIT recovery or quarantine behavior.",
      "scope": "R1 qualification subset of #16. One Mac, one org/private repository, first reviewed execution backend and maximum one active worker. A passing happy-path job is not evidence for lost ACK/acquisition/JIT recovery.",
      "tdd": [
        "Cover #1 required ACK, acquisition and JIT loss boundaries and idle-assignment/drain race with real evidence.",
        "Demonstrate no busy-job termination under normal scale-down, no duplicate execution, no non-owned cleanup.",
        "List every skipped case as unresolved."
      ],
      "acceptance": [
        "Exact reviewed commit, pinned identities, private runner-group/repository policy and bounded concrete maintainer authorization recorded before dispatch.",
        "Real hosted control-plane/private canary workflow executes one owned worker with independently observed terminal/absence evidence.",
        "Keep R1 qualification open until all required cases pass or reach the specified safe quarantine outcome.",
        "Broader #16 remains open."
      ],
      "risk": "High",
      "kind": "gate",
      "priority": "P0",
      "status": "Blocked",
      "test_profile": [
        "offline",
        "trusted-runtime",
        "trusted-live-github"
      ],
      "github_issue": {
        "number": 69,
        "url": "https://github.com/1XP-AI/gh-runnerd/issues/69"
      }
    }
  ]
}