{
  "id": "page:docs-v6-spec-and-roadmap-v6-1-v6-1-priorities",
  "_kind": "Page",
  "_file": "wiki/docs/v6-spec-and-roadmap/v6-1/v6-1-priorities.md",
  "_cluster": "wiki",
  "attributes": {
    "nodeKind": "Page",
    "sourcePath": "docs/v6-spec-and-roadmap/v6.1/v6.1-priorities.md",
    "sourceKind": "repo-docs",
    "title": "v6.1 Priorities — Actionable Work Items",
    "displayName": "v6.1 Priorities — Actionable Work Items",
    "slug": "docs/v6-spec-and-roadmap/v6-1/v6-1-priorities",
    "articlePath": "wiki/docs/v6-spec-and-roadmap/v6.1/v6.1-priorities.md",
    "article": "\n# v6.1 Priorities — Actionable Work Items\n\nWork items derived from the gap inventory, ordered by impact and feasibility.\n\n## P0: Fix Before Next Release\n\nThese are active regressions or blockers that prevent stable CI/CD.\n\n| Item | Gap ID | Layer | Description | Status |\n|------|--------|-------|-------------|--------|\n| Codex+google websocket proxy bypass | G-L3-02 | L3 | Codex websocket connections bypass OPENAI_BASE_URL (#200) | Open |\n| hooks-adapter-cli version hash inconsistency | G-X-04 | Cross | 12-char vs variable-length SHA in staging versions causes ETARGET | Fixed (publish.yml) |\n| Windows npm spawn | G-X-03 | Cross | `shell: true` needed on Windows for npm.cmd | Fixed |\n| ESM process module loading | G-L8-01 | L8 | NODE_PATH hack for ESM import resolution (#196) | Fixed (workaround) |\n| Process code drift | G-L13-02 | L13 | Stable invocation keys for replay (#169, PR #178) | In review |\n\n## P1: Architectural Improvements\n\nThese address S2 (Major) gaps that cause ongoing friction.\n\n### 1. Unified Provider Abstraction (G-L2-01, G-L2-02)\n\n**Problem:** Provider routing is split across transport-adapter (proxy), adapters-cli (translation), and agent-catalog (metadata).\n\n**Proposed:** Extract a `ProviderResolver` interface that:\n- Resolves provider + model → transport config + credentials\n- Lives in agent-catalog (metadata-backed) or a new thin package\n- Used by transport-adapter proxy and adapters-cli launch\n\n**Effort:** Medium. Seam already exists in agent-catalog's `getProviderTranslation()`.\n\n### 2. Explicit Workspace Contract (G-L7-01)\n\n**Problem:** The workspace concept (what the agent sees/modifies) is implicit in `.a5c/runs/` layout.\n\n**Proposed:** Define a `WorkspaceContract` interface in babysitter-sdk:\n- `resolveWorkspace(runDir)` → `{ root, processDir, artifactsDir, runsDir }`\n- Used by ProcessContext, task execution, and plugin install\n- Maps to graph's Workspace and Worktree node kinds\n\n**Effort:** Small. Most paths already exist in scattered utilities.\n\n### 3. Transport Codec Completion (G-L3-01)\n\n**Problem:** Codec architecture (tool schema translation, cost normalization) was planned but partially implemented.\n\n**Proposed:** Complete the codec refactor per the existing plan:\n- TransportCodec interface per protocol (anthropic, openai-chat, google, bedrock)\n- Tool schema translation between formats\n- Cost/usage normalization across providers\n- Already has a plan document in the repo\n\n**Effort:** Large. Significant API surface across multiple protocols.\n\n### 4. Unified Governance API (G-L14-01)\n\n**Problem:** Governance concerns scattered across 3 packages.\n\n**Proposed:** Define a `GovernancePolicy` interface in babysitter-sdk:\n- Approval routing (delegates to tasks-adapter)\n- Evidence claims (delegates to atlas)\n- Trust level evaluation (backed by graph TrustLevel records)\n\n**Effort:** Medium. Integration layer, not new functionality.\n\n### 5. Graph Cluster Alignment (G-X-01)\n\n**Problem:** Some node kinds are in wrong clusters (ProviderTranslation in \"extensions\", TransportRuntime in \"extensions\").\n\n**Proposed:** Move graph YAML files to align node kinds with their architectural layer clusters:\n- ProviderTranslation → compute cluster\n- TransportRuntime → compute cluster\n- AdapterModel → capabilities-and-models cluster\n\n**Effort:** Small. YAML file moves + index rebuild.\n\n## P2: Package Metadata\n\nLow-effort improvements that help navigation.\n\n| Item | Gap ID | Effort | Description |\n|------|--------|--------|-------------|\n| Add layer tags to package.json | G-X-02 | Small | Add `\"atlas\": { \"layers\": [\"L13\"] }` to each package.json |\n| Package README layer references | G-X-02 | Small | Each README should state which layer(s) the package implements |\n| Graph node kind references in exports | G-X-02 | Small | TypeScript exports should reference their graph node kinds in JSDoc |\n\n## P3: Future Architecture (Deferred from v6.0)\n\nThese are explicitly deferred. Tracked here for completeness.\n\n| Item | Gap IDs | Layer | Notes |\n|------|---------|-------|-------|\n| Durable agent memory | G-L12-01, G-L12-02 | L12 | Requires vector store, embedding models, retrieval pipeline |\n| Filesystem/network sandbox | G-L9-01, G-L9-02, G-L9-03 | L9 | Requires runtime policy enforcement engine |\n| Model selection engine | G-L1-01 | L1 | Requires capability negotiation protocol |\n| Team agent coordination | G-L6-01 | L6 | Beyond subagent dispatch — shared context, handoff |\n| Marketplace identity/billing | G-L6-02 | L6 | Enterprise feature |\n| NIST/OWASP compliance mapping | G-L14-02 | L14 | Enterprise governance |\n\n## Decision Criteria\n\nWork items should be prioritized based on:\n\n1. **User impact:** Does this gap cause visible failures or confusion?\n2. **Agent impact:** Does this gap prevent autonomous agent workflows?\n3. **Graph alignment:** Does fixing this bring packages closer to the graph source of truth?\n4. **Incremental value:** Can this be shipped independently without blocking other work?\n\nP0 items ship immediately. P1 items are v6.1 scope. P2 items are opportunistic. P3 items are tracked but not planned.\n",
    "documents": []
  },
  "outgoingEdges": [],
  "incomingEdges": [
    {
      "from": "page:docs-v6-spec-and-roadmap-v6-1",
      "to": "page:docs-v6-spec-and-roadmap-v6-1-v6-1-priorities",
      "kind": "contains_page"
    }
  ]
}