@openlife/cli 1.7.4 → 1.7.6

package/docs/audit/DOC-PARITY-GAPS.md

@@ -1,351 +0,0 @@
-# Doc/Code Parity Gaps — openlife-cli
-
-**Source of truth:** `docs/audit/CLI-MATRIX.md` (156 `.command()` registrations; 125 leaf commands across 11 top-level standalones + 34 groups + 2 subgroups)
-**Docs cross-checked:** `INSTALL.md` (139 lines), `README.md` (137 lines), `CLAUDE.md` (467 lines, focus on "Build / Run / Test")
-**Generated:** 2026-05-11
-
----
-
-## Summary
-
-| Severity | Count |
-|----------|-------|
-| DRIFT-CRITICAL (doc says X, code doesn't) | 3 |
-| DRIFT-HIGH (in code, undocumented anywhere) | 105 |
-| DRIFT-MEDIUM (in some docs but not others) | 12 |
-| DRIFT-LOW (description/default value drift) | 4 |
-| **Total gaps** | **124** |
-
-Out of 125 leaf commands in the matrix, **only 20 are mentioned in any of the three docs**. The vast majority of the CLI surface is undocumented. The three critical drift items are doc-side: docs reference commands that do not exist in code (`mode`, `chat`).
-
----
-
-## DRIFT-CRITICAL
-
-Docs reference commands or flag combinations that **do not exist** in `src/index.ts`. Users following the docs will see "unknown command".
-
-| # | Doc reference | What the doc says | What code actually does | Fix recommendation |
-|---|--------------|-------------------|--------------------------|--------------------|
-| 1 | `INSTALL.md` line 66 ("Primeiro uso") | `openlife mode set framework --host claude-code` | No top-level `mode` command exists. There is **no** `mode` group anywhere in `src/index.ts`. Closest equivalent: `system install --profile framework --host claude-code` (line 467) or `system setup --profile framework --host claude-code` (line 505). | Replace with `openlife system setup --profile framework --host claude-code` (already used on `INSTALL.md` line 11) **or** add a `mode` command alias in code. |
-| 2 | `INSTALL.md` line 129 ("Diagnóstico") | `openlife mode status` | No `mode` command. No `mode status` either. Likely intended: `system status` or `models status`. | Replace with `openlife system status` (already on line 128) or remove the line. |
-| 3 | `README.md` line 65 ("Chat local") | `openlife chat` | No top-level `chat` command exists. The closest is `aiobuilder chat <mensagem...>` (line 1845) which is scoped to design-mode intents only — not a general chat surface. `ask <mensagem...>` (line 399) is the canonical "send to Gatekeeper" entry but is not mentioned in README. | Replace with `openlife ask "<message>"` (semantically the chat-equivalent), or document `aiobuilder chat <message>` if the intent was design-mode only. |
-
----
-
-## DRIFT-HIGH
-
-Leaf commands present in `src/index.ts` (per CLI-MATRIX.md) that are **not mentioned in INSTALL.md, README.md, or CLAUDE.md**.
-
-Grouped by command group. Count: 105.
-
-### Top-level standalone commands (3 of 11 undocumented)
-| # | Command | Where it lives | Why it should be documented |
-|---|---------|----------------|------------------------------|
-| 1 | `restart` | src/index.ts:1171 | Operator-facing daemon control; users need it for autonomous mode |
-| 2 | `update` | src/index.ts:1186 | Dev-mode self-update; should be mentioned in INSTALL §9 (PATH/postinstall) |
-| 3 | `dream <entry...>` (alias `/dream`) | src/index.ts:1345 | User-facing personal journaling command; visible in `--help` |
-
-### `auth` group (2/2 undocumented)
-| # | Command | Where | Why |
-|---|---------|-------|-----|
-| 4 | `auth gemini` | src/index.ts:246 | Required for Gemini CLI executor; INSTALL §1 lists Gemini as optional but never shows the OAuth command |
-| 5 | `auth openai` (alias `codex`) | src/index.ts:259 | Required for Codex CLI executor; same omission |
-
-### `models` group (5/6 undocumented; `models status` mentioned in README & INSTALL)
-| # | Command | Where | Why |
-|---|---------|-------|-----|
-| 6 | `models set <provider/model>` | 323 | Primary model swap; users hit this whenever they change provider |
-| 7 | `models fallbacks list` | 339 | Fallback chain inspection |
-| 8 | `models fallbacks add <provider/model>` | 348 | Fallback chain mutation |
-| 9 | `models fallbacks remove <provider/model>` | 369 | Fallback chain mutation |
-| 10 | `models fallbacks clear` | 384 | Fallback chain reset |
-
-### `system` group (3/7 undocumented; `system setup`, `system doctor`, `system status` mentioned)
-| # | Command | Where | Why |
-|---|---------|-------|-----|
-| 11 | `system install` | 467 | Alias of `system setup` per matrix description, but docs only show `system setup` |
-| 12 | `system init-project` | 552 | Creates `OPENLIFE_PROJECT.md` — first-day onboarding step |
-| 13 | `system doctor-world` | 560 | World-class diagnostic |
-| 14 | `system bootstrap-world` | 567 | World-class bootstrap |
-
-### `memory` group (5/5 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 15 | `memory status` | 573 |
-| 16 | `memory provider-list` | 577 |
-| 17 | `memory provider-use <name>` | 581 |
-| 18 | `memory test <name>` | 585 |
-| 19 | `memory search <query...>` | 589 |
-
-### `swarm` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 20 | `swarm run [goal...]` | 600 |
-| 21 | `swarm consensus-test` | 621 |
-
-### `governance` group (5/5 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 22 | `governance status` | 627 |
-| 23 | `governance audit` | 631 |
-| 24 | `governance risk-check <goal>` | 635 |
-| 25 | `governance consent <scope> <userId>` | 639 |
-| 26 | `governance policy show` | 644 |
-
-### `benchmark` group (3/3 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 27 | `benchmark run` | 669 |
-| 28 | `benchmark compare` | 673 |
-| 29 | `benchmark report` | 677 |
-
-### `agents` group (2/3 undocumented; `agents list` in README only)
-| # | Command | Where |
-|---|---------|-------|
-| 30 | `agents create <id>` | 692 |
-| 31 | `agents show <id>` | 703 |
-
-### `squads` group (1/2 undocumented; `squads list` in README only)
-| # | Command | Where |
-|---|---------|-------|
-| 32 | `squads create <id>` | 747 |
-
-### `skills` group (1/2 undocumented; `skills list` in README only)
-| # | Command | Where |
-|---|---------|-------|
-| 33 | `skills create <id>` | 766 |
-
-### `route` group (1/1 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 34 | `route intent <goal...>` | 869 |
-
-### `mcp` group (1/2 undocumented; `mcp status --real` in README only)
-| # | Command | Where |
-|---|---------|-------|
-| 35 | `mcp create <id>` | 900 |
-
-### `catalog` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 36 | `catalog doctor` | 933 |
-| 37 | `catalog list` | 952 |
-
-### `teams` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 38 | `teams list` | 958 |
-| 39 | `teams upsert <id>` | 969 |
-
-### `networks` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 40 | `networks list` | 986 |
-| 41 | `networks upsert <id>` | 996 |
-
-### `media` group (4/4 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 42 | `media route` | 1013 |
-| 43 | `media route-status` | 1027 |
-| 44 | `media doctor` | 1061 |
-| 45 | `media read-image <imagePath>` | 1083 |
-
-### `orchestration` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 46 | `orchestration contract` | 1126 |
-| 47 | `orchestration snapshot` | 1130 |
-
-### `smoke` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 48 | `smoke railway` | 1324 |
-| 49 | `smoke telegram` | 1331 |
-
-### `service` group (6/6 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 50 | `service list` | 1358 |
-| 51 | `service create <serviceId>` | 1367 |
-| 52 | `service status <serviceId>` | 1380 |
-| 53 | `service pause <serviceId>` | 1393 |
-| 54 | `service resume <serviceId>` | 1406 |
-| 55 | `service events <serviceId>` | 1419 |
-
-### `task` group (3/3 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 56 | `task list` | 1434 |
-| 57 | `task run <goal...>` | 1444 |
-| 58 | `task status <taskId>` | 1457 |
-
-### `job` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 59 | `job list` | 1470 |
-| 60 | `job events <jobId>` | 1477 |
-
-### `runtime` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 61 | `runtime probe` | 1483 |
-| 62 | `runtime list` | 1487 |
-
-### `reversa` group (6/6 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 63 | `reversa status` | 1493 |
-| 64 | `reversa next` | 1499 |
-| 65 | `reversa mode` | 1506 |
-| 66 | `reversa run-phase <phase>` | 1516 |
-| 67 | `reversa run-all` | 1522 |
-| 68 | `reversa export-json` | 1526 |
-
-### `sources` group (4/4 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 69 | `sources list` | 1532 |
-| 70 | `sources guard-check <url>` | 1536 |
-| 71 | `sources scaffold <type> <id>` | 1543 |
-| 72 | `sources import-ref <type> <id> <url>` | 1559 |
-
-### `teammate` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 73 | `teammate list` | 1579 |
-| 74 | `teammate upsert <id> <owner> <status> <title>` | 1586 |
-
-### `learning` group (2/2 undocumented) — distinct from `learn` group
-| # | Command | Where |
-|---|---------|-------|
-| 75 | `learning add <taskId> <summary>` | 1593 |
-| 76 | `learning list` | 1597 |
-
-### `context` group (3/3 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 77 | `context init` | 1605 |
-| 78 | `context doctor` | 1608 |
-| 79 | `context summary` | 1613 |
-
-### `plugin` group (5/5 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 80 | `plugin install <id>` | 1626 |
-| 81 | `plugin list` | 1629 |
-| 82 | `plugin inspect <id>` | 1632 |
-| 83 | `plugin disable <id>` | 1635 |
-| 84 | `plugin enable <id>` | 1638 |
-
-### `pilot` group (4/4 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 85 | `pilot create <id>` | 1646 |
-| 86 | `pilot list` | 1649 |
-| 87 | `pilot report <id>` | 1652 |
-| 88 | `pilot complete <id>` | 1657 |
-
-### `eval` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 89 | `eval judge` | 1665 |
-| 90 | `eval report <mission>` | 1668 |
-
-### `learn` group (2/2 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 91 | `learn from-last-run` | 1676 |
-| 92 | `learn list` | 1679 |
-
-### `enterprise` group (1/1 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 93 | `enterprise doctor` | 1684 |
-
-### `designmd` group (5/5 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 94 | `designmd status` | 1691 |
-| 95 | `designmd import` | 1698 |
-| 96 | `designmd list` | 1711 |
-| 97 | `designmd show <profileId>` | 1715 |
-| 98 | `designmd apply <profileId> <source> <title> <designPath>` | 1720 |
-
-### `aiobuilder` group (6/6 undocumented)
-| # | Command | Where |
-|---|---------|-------|
-| 99 | `aiobuilder mode` | 1730 |
-| 100 | `aiobuilder create-agent <id>` | 1741 |
-| 101 | `aiobuilder create-squad <id>` | 1749 |
-| 102 | `aiobuilder validate-catalog` | 1754 |
-| 103 | `aiobuilder generate-ui <featureName>` | 1763 |
-| 104 | `aiobuilder chat <mensagem...>` | 1845 |
-
-### `ask` (1/1 — top-level, not mentioned in any doc by name)
-| # | Command | Where |
-|---|---------|-------|
-| 105 | `ask <mensagem...>` | 399 |
-
-Note: `ask` is the canonical user-facing entry to the Gatekeeper and is heavily exercised in tests, yet is referenced in **none** of the three docs. README's `openlife chat` (drift-critical #3) is the closest doc-side intent and should be replaced by `openlife ask`.
-
----
-
-## DRIFT-MEDIUM
-
-Commands that **are** documented in at least one doc but missing from another where they would be relevant.
-
-| # | Command | In README? | In INSTALL? | In CLAUDE? | Gap description |
-|---|---------|-----------|-------------|-----------|------------------|
-| 1 | `install` (top-level) | YES (line 31) | NO | NO | INSTALL.md has its own bash recipe and mentions `npm install -g`; the actual `openlife install` interactive command isn't mentioned in INSTALL — paradoxical |
-| 2 | `up` | YES (line 53) | NO | NO | One-shot daily-driver command. Should appear in INSTALL §6 or §10 |
-| 3 | `start --daemon` | YES (line 59) | NO | NO | Manual daemon start. INSTALL §6 talks about autonomous mode but uses `agent start` instead of `start --daemon` |
-| 4 | `status` (top-level) | YES (line 82, 128) | NO | NO | INSTALL §10 (Diagnóstico) lists `system status` and `mode status` but not the top-level `status` |
-| 5 | `doctor` (top-level) | YES (line 83, 126) | NO | NO | INSTALL §10 lists `system doctor` only, not the universal top-level `doctor` |
-| 6 | `phase1-check` | YES (line 85, 129) | NO | NO | INSTALL never mentions phase1-check |
-| 7 | `agents list` | YES (line 86) | NO | NO | INSTALL has no command examples beyond `models status`/`system doctor`/etc. |
-| 8 | `squads list` | YES (line 87) | NO | NO | Same gap as agents list |
-| 9 | `skills list` | YES (line 88) | NO | NO | Same |
-| 10 | `mcp status --real` | YES (line 89) | NO | NO | Same |
-| 11 | `system setup` (full form) | YES (line 41, 47) | YES (line 11) | NO | CLAUDE.md "Build / Run / Test" never shows the install/setup commands — only `npm run build` and `node bin/openlife.js --help` |
-| 12 | `agent install`/`agent start`/`agent status` | NO | YES (lines 17, 23, 82-83) | NO | INSTALL uses `openlife agent` alias; README uses `start --daemon` instead — the two docs disagree on which surface to teach |
-
----
-
-## DRIFT-LOW
-
-Description/default-value drift. Short paragraphs, no table needed.
-
-1. **`install --host` default vs INSTALL recipe.** Matrix line 18 shows `--host` defaults to `claude-code`. INSTALL.md hard-codes `--host claude-code` on the recipes, which is fine, but README mirrors the same hard-coding; neither doc states that the flag is optional and defaults to `claude-code`. Low-impact but a documentation polish item.
-
-2. **`system setup` description.** The matrix records `system setup` as "Alias de system install para onboarding completo" (line 505). INSTALL.md (line 11) and README.md (lines 41, 47) all surface `system setup` as the canonical command; **`system install` is never mentioned in docs at all** (already captured as DRIFT-HIGH #11). Calling `system setup` an alias of an undocumented command is confusing — recommend updating the description in `src/index.ts` to drop the "alias" language or swap which command is documented.
-
-3. **`media route --prefer-gemini` default.** Default is `'true'` (string) per matrix line 1013. Not documented anywhere, so users can't know the default. Minor, since the only references to media routing in docs are the `OPENLIFE_PREFER_GEMINI_FOR_MEDIA` env var in CLAUDE.md (line 184).
-
-4. **`dream` alias `/dream`.** The Telegram-style alias is registered (matrix line 1345) but neither README nor INSTALL mentions `dream` or `/dream`. CLAUDE.md mentions it only in the dev-time test fixture context. Minor — but the alias is unusual enough (starts with `/`) that documentation would help.
-
----
-
-## Notes
-
-- **Commands clearly internal/diagnostic** — these may not warrant prominent doc placement but should still appear in a comprehensive `docs/commands.md` (referenced by INSTALL §11 but not present in repo):
-  - `phase1-check` (Phase-1 runtime check)
-  - `smoke railway`, `smoke telegram` (CI smokes)
-  - `enterprise doctor`, `media doctor`, `catalog doctor`, `system doctor-world`, `governance audit`
-  - `orchestration contract`, `orchestration snapshot`
-  - `runtime probe`, `runtime list`
-  - `swarm consensus-test`, `benchmark run/compare/report`
-
-- **Narrative-only references** in the three docs that are *not* drift, just descriptive prose (correctly excluded from counts above):
-  - INSTALL "Bundle automatizado (local/WSL/Linux)" referring to a shell script, not a CLI command
-  - README "Telegram" section describing routing behavior (no CLI surface invoked)
-  - CLAUDE.md "Test suite — important pattern" section showing `npm run test:*` (npm scripts, not CLI commands)
-  - CLAUDE.md "Build / Run / Test" mentions `npm run admin` → admin server, not a CLI command
-
-- **Linked docs not yet present**: INSTALL.md §11 links to `docs/quickstart.md`, `docs/commands.md`, `docs/autonomous-install.md` — only `docs/autonomous-install.md` is confirmed to exist (`test_distribution_installability.ts` line 21). The other two would be the natural homes for the 105 DRIFT-HIGH commands and were not scored here as gaps because they are missing files, not drift.
-
-- **Doc-only doctor-set bias.** Both INSTALL and README repeat similar small diagnostic command sets (`status`, `doctor`, `system doctor`, `models status`) rather than steering users to the broader `catalog doctor`, `media doctor`, `enterprise doctor`, `runtime probe` surface. Worth a future doc consolidation.
-
-- **`agent <verb>` is an alias bridge.** Top-level `agent <verb>` (line 1248, "Alias compatível com documentação: agent install|start|status → autonomous lifecycle") is *deliberately* documented in INSTALL.md even though the underlying lifecycle command structure is different. No drift to flag.
-
-- **CLAUDE.md scope.** The brief asked to focus on CLAUDE.md's "Build / Run / Test" section. That section references only build/test scripts (`npm install`, `npm run build`, `node bin/openlife.js --help`, `npm run dev`, `npm run admin`, `npm run test:*`). No `openlife <subcommand>` invocations outside `--help` appear, so CLAUDE.md effectively contributes no command-level documentation coverage and is the most under-documented of the three docs from a CLI-surface standpoint.

package/docs/audit/ORCHESTRATOR-MATRIX.md

@@ -1,136 +0,0 @@
-# Orchestrator Matrix — openlife-cli
-
-**Source:** `src/orchestrator/` (79 .ts files + `providers/` with 12 .ts files = 91 total)
-**Total classes:** 91 (82 core + 9 in `providers/`)
-**Generated:** 2026-05-11
-
-Phase 1 of Total Feature Audit. Captures public surface area (constructor + non-`private`/non-`protected` methods) for every class. Subsequent phases will derive unit tests from this matrix.
-
-## Core Classes
-
-| # | File | Class | Constructor | Public Methods | Collaborators | Responsibility |
-|---|------|-------|-------------|----------------|---------------|----------------|
-| 1 | `AgentRegistry.ts` | `AgentRegistry` | `(private provider: AgentProvider = new CompositeAgentProvider( new CloudAgentProvider(process.env....)` | `all(): RegisteredAgent[]`<br>`findByCapability(capability): RegisteredAgent[]`<br>`registerTemporary(agent)`<br>`findMiniSquad(capabilities): RegisteredAgent[]`<br>`findPrioritySet(): RegisteredAgent[]` | `AgentProvider`, `FileAgentProvider`, `CloudAgentProvider`, `CompositeAgentProvider` | Asset registry for agent |
-| 2 | `AgentScoring.ts` | `AgentScoring` | `()` | `getAll(): Record<string, AgentScore>`<br>`record(agentId, status)` | — | Tracks agent scores |
-| 3 | `AgentTeam.ts` | `AgentTeamRegistry` | `(seed?: AgentTeam[])` | `list(): AgentTeam[]`<br>`upsert(team): void`<br>`resolveByDomain(domain): AgentTeam \| null` | — | Asset registry for agentteam |
-| 4 | `ArbitrationAgent.ts` | `ArbitrationAgent` | `(private brain: Brain)` | `arbitrate(goal, branches): Promise<ArbitrationResult>` | `Brain` | Arbitration agent (LLM-backed) |
-| 5 | `ArbitrationScorecard.ts` | `ArbitrationScorecard` | `()` | `score(branches): BranchScore[]` | — | Scores/ranks arbitration |
-| 6 | `AssetPromotionEngine.ts` | `AssetPromotionEngine` | `()` | `promote(candidate): PromotedAsset` | `PromotionCandidate` | AssetPromotion engine |
-| 7 | `AssetReuseRouter.ts` | `AssetReuseRouter` | `()` | `find(goal): ReusedAsset[]` | — | Routes assetreuse requests to handlers |
-| 8 | `BenchmarkEngine.ts` | `BenchmarkEngine` | `()` | `record(sample)`<br>`compare(benchmarkId)`<br>`getAll(): BenchmarkSample[]` | — | Benchmark engine |
-| 9 | `Brain.ts` | `Brain` | `()` | `think(systemPrompt, userMessage): Promise<string>`<br>`formatProviderError(provider, model, error, opts): Error` | `ModelManager` | Brain component |
-| 10 | `CadenceEngine.ts` | `CadenceEngine` | `()` | `schedule(kind, title, ownerRole, linkedMissionId): CadenceEvent`<br>`updateStatus(id, status)`<br>`getAll(): CadenceEvent[]` | — | Cadence engine |
-| 11 | `CapabilityRouter.ts` | `CapabilityRouter` | `(private registry: AgentRegistry)` | `route(task, input): CapabilityRoute` | `RegisteredAgent`, `AgentRegistry`, `ClassifiedTask`, `TaskIntent` | Routes capability requests to handlers |
-| 12 | `CommandLanguage.ts` | `CommandLanguage` | `()` | `parse(input): ParsedCommand` | — | Parses command language |
-| 13 | `CommandRouter.ts` | `CommandRouter` | `()` | `route(parsed): string \| null` | `ParsedCommand`, `AgentRegistry`, `SquadRegistry`, `SkillRegistryV2`, `OrgStructure` | Routes command requests to handlers |
-| 14 | `ConsequenceForecaster.ts` | `ConsequenceForecaster` | `()` | `forecast(goal, governanceRisk, basePolicy, executors): ConsequenceForecast` | `ExecutionPolicyDecision`, `ExecutorKind` | Forecasts consequence |
-| 15 | `DynamicAgentBuilder.ts` | `DynamicAgentBuilder` | `(private brain: Brain)` | `propose(goal, existingAgents): Promise<DynamicAgentSpec \| null>` | `Brain`, `RegisteredAgent` | Builds dynamicagent |
-| 16 | `DynamicAgentExecutor.ts` | `DynamicAgentExecutor` | `(private taskExecutor: TaskExecutor)` | `execute(agent, goal, taskId): Promise<string>` | `TaskExecutor`, `DynamicAgentSpec` | Executes dynamicagent |
-| 17 | `EnterpriseAgenticCore.ts` | `EnterpriseDoctor` | `(private root = process.cwd())` | `run()` | — | Diagnoses Enterprise health |
-| 18 | `ExecutionBoard.ts` | `ExecutionBoard` | `()` | `add(item)`<br>`updateStatus(id, status, blockerReason)`<br>`updateDependencies(id, dependencies)`<br>`getAll(): BoardItem[]` | — | Execution board / list state |
-| 19 | `ExecutionModePolicy.ts` | `ExecutionModePolicy` | `()` | `decide(task, input): ExecutionPolicyDecision` | `ClassifiedTask`, `TaskIntent` | Policy decisions for executionmode |
-| 20 | `ExecutionRouter.ts` | `ExecutionRouter` | `()` | `route(intent): RoutedExecutionPath` | `ExecutionIntent` | Routes execution requests to handlers |
-| 21 | `ExecutorHealth.ts` | `ExecutorHealth` | `()` | `getAll(): Record<string, ExecutorHealthState>` | `ExecutorKind` | ExecutorHealth component |
-| 22 | `ExternalCatalogRegistry.ts` | `ExternalCatalogRegistry` | `()` | `list(): ExternalCatalogSource[]` | — | Asset registry for externalcatalog |
-| 23 | `Gatekeeper.ts` | `Gatekeeper` | `()` | `routeTask(task, userInput, userId, options): Promise<string>` | `TaskIntent`, `ClassifiedTask`, `OmniMemory`, `SquadManager`, `SkillRegistry` | Gatekeeper component |
-| 24 | `Gateway.ts` | `Gateway` | `()` | `start()`<br>`shutdown(reason): Promise<void>`<br>`processTextForTest(userId, text): Promise<string>`<br>`processImageForTest(userId, imagePath, caption): Promise<string>` | `IntentClassifier`, `Gatekeeper`, `VoiceManager`, `GovernanceLayer` | Gateway component |
-| 25 | `GovernanceConsentStore.ts` | `GovernanceConsentStore` | `(private root: string = process.cwd())` | `approve(userId, scope)`<br>`hasRecentApproval(userId, scope, windowMinutes): boolean` | — | Persistent store for governanceconsent |
-| 26 | `GovernanceLayer.ts` | `GovernanceLayer` | `()` | `evaluate(goal, project): GovernanceDecision`<br>`redact(text): string` | `GovernancePolicyStore` | Governance layer |
-| 27 | `GovernancePolicyStore.ts` | `GovernancePolicyStore` | `(private root: string = process.cwd())` | `load(project)` | — | Persistent store for governancepolicy |
-| 28 | `GovernanceScopePolicy.ts` | `GovernanceScopePolicy` | `()` | `decide(task, input, project): GovernanceScopeDecision`<br>`filterCapabilities(capabilities, scope)` | `ClassifiedTask`, `TaskIntent`, `GovernancePolicyStore` | Policy decisions for governancescope |
-| 29 | `EnterpriseAgenticCore.ts` | `InstitutionalContextStore` | `(private root = process.cwd())` | `requiredFiles()`<br>`init(name)`<br>`doctor()`<br>`summary()` | — | Persistent store for institutionalcontext |
-| 30 | `IntentClassifier.ts` | `IntentClassifier` | `()` | `classify(userInput): Promise<ClassifiedTask>` | — | O Classificador Cognitivo do OPEN-LIFE. |
-| 31 | `JobLifecycle.ts` | `JobLifecycleStore` | `(workspaceId = 'default')` | `enqueue(goal, workspaceId, serviceId, taskId): JobRecord`<br>`transition(jobId, workspaceId, status, summary, ownerAgentId, blocker): JobRecord \| null`<br>`list(limit): JobRecord[]`<br>`events(jobId): JobEvent[]` | — | Persistent store for joblifecycle |
-| 32 | `EnterpriseAgenticCore.ts` | `LearningLoopStore` | `(private root = process.cwd())` | `fromLastRun(mission, feedback)`<br>`list()` | — | Persistent store for learningloop |
-| 33 | `LearningRouter.ts` | `LearningRouter` | `(private capabilityRouter: CapabilityRouter, private scoring: AgentScoring)` | `route(task, input): CapabilityRoute` | `CapabilityRouter`, `CapabilityRoute`, `AgentScoring`, `PerformanceScorecard` | Routes learning requests to handlers |
-| 34 | `MediaManager.ts` | `MediaManager` | `()` | `extractAudio(inputPath, outputFilename): Promise<string>`<br>`cutVideo(inputPath, startTime, duration, outputFilename): Promise<string>` | — | Manages media |
-| 35 | `MemoryCuratorAgent.ts` | `MemoryCuratorAgent` | `(private brain: Brain)` | `curate(state): Promise<string[]>` | `Brain`, `MissionState` | MemoryCurator agent (LLM-backed) |
-| 36 | `EnterpriseAgenticCore.ts` | `MissionEvaluationStore` | `(private root = process.cwd())` | `judge(mission, criteriaCsv)`<br>`report(mission)`<br>`list()` | — | Persistent store for missionevaluation |
-| 37 | `MissionState.ts` | `MissionStateStore` | `()` | `persist(state): string` | `ExecutionState`, `RegisteredAgent`, `RegisteredSquad`, `RegisteredSkill` | Persistent store for missionstate |
-| 38 | `ModelManager.ts` | `ModelManager` | `()` | `getModelConfig(): ModelConfig`<br>`saveConfig(config)`<br>`parseModelString(input): ModelIdentifier` | — | Manages model |
-| 39 | `OperatingSystem.ts` | `OperatingSystem` | `()` | `decide(missionId): OperatingDecision`<br>`scheduleWeeklyCadence(anchor)` | `ExecutionBoard`, `BoardItem`, `RoleHandoff`, `CadenceEngine` | OperatingSystem component |
-| 40 | `OperationalMemoryStore.ts` | `OperationalMemoryStore` | `()` | `add(entry, namespace)`<br>`getAll(): OperationalMemoryEntry[]` | `LocalMemoryProvider`, `MemoryNamespace`, `MemoryRecord`, `MemoryOrchestrator` | Persistent store for operationalmemory |
-| 41 | `OptimizationLoop.ts` | `OptimizationLoop` | `()` | `generateSuggestions(): OptimizationSuggestion[]` | `PerformanceScorecard` | Optimization loop runner |
-| 42 | `OrchestrationLoop.ts` | `OrchestrationLoop` | `(private brain: Brain, private taskExecutor: TaskExecutor)` | `run(task, input, userId, recentHistory, intent): Promise<` | `Brain`, `TaskExecutor`, `ClassifiedTask`, `TaskIntent`, `createExecutionState` | Orchestration loop runner |
-| 43 | `OrgStructure.ts` | `OrgStructure` | `()` | `all(): OrgRole[]` | — | OrgStructure component |
-| 44 | `OutcomeSimulator.ts` | `OutcomeSimulator` | `()` | `simulate(input): OutcomeSimulationResult` | `ExecutorKind`, `ExecutionPolicyDecision` | Simulates outcome |
-| 45 | `ParallelOrchestrationLoop.ts` | `ParallelOrchestrationLoop` | `(private brain: Brain, private taskExecutor: TaskExecutor)` | `runResearchBranches(goal, taskId, maxBranches): Promise<ParallelRunResult>` | `Brain`, `TaskExecutor`, `ArbitrationAgent`, `ArbitrationResult`, `ArbitrationScorecard` | ParallelOrchestration loop runner |
-| 46 | `PerformanceScorecard.ts` | `PerformanceScorecard` | `()` | `record(entry)`<br>`getAll(): Record<string, PerformanceAggregate>` | — | Scores/ranks performance |
-| 47 | `EnterpriseAgenticCore.ts` | `PilotStore` | `(private root = process.cwd())` | `create(id, metric, service)`<br>`list()`<br>`report(id)`<br>`complete(id, outcome)` | — | Persistent store for pilot |
-| 48 | `PlannerAgent.ts` | `PlannerAgent` | `(private brain: Brain)` | `plan(input, recentHistory): Promise<PlannerOutput>` | `Brain` | Planner agent (LLM-backed) |
-| 49 | `EnterpriseAgenticCore.ts` | `PluginManifestStore` | `(private root = process.cwd())` | `riskTemplate(risk)`<br>`install(id, capabilities, risk)`<br>`list()`<br>`inspect(id)`<br>`setStatus(id, status)` | — | Persistent store for pluginmanifest |
-| 50 | `PromotionPipeline.ts` | `PromotionPipeline` | `()` | `add(candidate, context)`<br>`getAll(): PromotionCandidate[]`<br>`clear()` | `MemoryNamespacePolicy`, `MemoryPromotionFlow` | Promotion pipeline |
-| 51 | `PromotionReviewGate.ts` | `PromotionReviewGate` | `()` | `approve(candidate): boolean` | `PromotionCandidate` | PromotionReview gate |
-| 52 | `ReleaseGate.ts` | `ReleaseGate` | `()` | `run(releaseId, missionId): Promise<ReleaseChecklist>` | `SystemDoctor`, `TestHarness`, `ReleaseWorkflow`, `ReleaseChecklist`, `OperatingSystem` | Release gate |
-| 53 | `ReleaseWorkflow.ts` | `ReleaseWorkflow` | `()` | `create(releaseId, checks): ReleaseChecklist`<br>`getAll(): ReleaseChecklist[]` | — | Release workflow |
-| 54 | `ReuseEngine.ts` | `ReuseEngine` | `()` | `findHints(goal): ReuseHint[]`<br>`enrichAgents(agents, hints): RegisteredAgent[]` | `RegisteredAgent` | Reuse engine |
-| 55 | `ReviewerAgent.ts` | `ReviewerAgent` | `(private brain: Brain)` | `review(goal, executorOutput, successCriteria): Promise<ReviewFinding>` | `Brain`, `ReviewFinding` | Reviewer agent (LLM-backed) |
-| 56 | `RoleHandoff.ts` | `RoleHandoff` | `()` | `create(input)`<br>`updateStatus(id, status)`<br>`getAll(): HandoffRecord[]` | — | RoleHandoff component |
-| 57 | `RuntimeHealthMonitor.ts` | `RuntimeHealthMonitor` | `(private root: string = process.cwd())` | `isCoolingDown(executor): boolean`<br>`markFailure(executor, reason, cooldownMinutes)`<br>`markHealthy(executor)`<br>`all(): Record<string, RuntimeCooldownState>` | `ExecutorKind` | Monitors runtimehealth |
-| 58 | `RuntimePolicy.ts` | `RuntimePolicy` | `()` | `decide(intent, explicitExecutors): RuntimePolicyDecision`<br>`recordResult(executor, ok, reason)`<br>`status()`<br>`saveSnapshot()` | `ExecutorKind`, `ExecutorHealth`, `RuntimeHealthMonitor` | Policy decisions for runtime |
-| 59 | `RuntimeProbe.ts` | `RuntimeProbe` | `()` | `probeAll(): Promise<RuntimeProbeResult[]>`<br>`probe(executor): Promise<RuntimeProbeResult>` | `ExecutorKind`, `ExecutorHealth`, `ExecutorHealthState` | Probes runtime health |
-| 60 | `RuntimeRegistry.ts` | `RuntimeRegistry` | `(workspaceId = 'default')` | `probe(): RuntimeEntry[]`<br>`list(): RuntimeEntry[]` | — | Asset registry for runtime |
-| 61 | `SandboxPolicy.ts` | `SandboxPolicy` | `()` | `evaluate(executor, riskLevel): SandboxDecision` | `ExecutorKind` | Policy decisions for sandbox |
-| 62 | `SecurityDownloadGuard.ts` | `SecurityDownloadGuard` | `()` | `validateUrl(url): GuardScanResult`<br>`scanFileList(paths): GuardScanResult`<br>`scanExtractedDir(rootDir): GuardScanResult` | — | Guards securitydownload |
-| 63 | `SecurityEventStore.ts` | `SecurityEventStore` | `(private root: string = process.cwd())` | `append(event)`<br>`exists(): boolean`<br>`path(): string` | `GovernanceAuditEvent` | Persistent store for securityevent |
-| 64 | `ServiceCompletionPolicy.ts` | `ServiceCompletionPolicy` | `()` | `enforceCriteria(criteria): string[]`<br>`evaluate(state): ServiceCompletionReport` | `ExecutionState` | Policy decisions for servicecompletion |
-| 65 | `ServiceState.ts` | `ServiceStateStore` | `(private workspaceId = 'default')` | `list(limit): ServiceState[]`<br>`upsertFromMission(state): string \| null`<br>`setStatus(serviceId, status, reason, taskId): ServiceState \| null`<br>`listEvents(serviceId, limit): ServiceEvent[]`<br>`promoteTaskToService(state, serviceId, objective): ServiceState` | `MissionState` | Persistent store for servicestate |
-| 66 | `SkillLearningLoop.ts` | `SkillLearningLoopStore` | `(private workspaceId = 'default')` | `add(sourceTaskId, summary): SkillLearningCandidate` | — | Persistent store for skilllearningloop |
-| 67 | `SkillManager.ts` | `SkillManager` | `(private cloudFile: string = process.env.OPENLIFE_CLOUD_SKILLS_FILE \|\| '.artifacts/cloud-skills.json')` | `create(skill)`<br>`patch(id, content)`<br>`setStatus(id, status)` | — | Manages skill |
-| 68 | `SkillNetwork.ts` | `SkillNetworkRegistry` | `(seed?: SkillNetwork[])` | `list(): SkillNetwork[]`<br>`upsert(network): void`<br>`resolveCapabilities(domain, goal): string[]` | — | Asset registry for skillnetwork |
-| 69 | `SkillRegistryV2.ts` | `SkillRegistryV2` | `(private provider: SkillProvider = new CompositeSkillProvider( new CloudSkillProvider(process.env....)` | `all(): RegisteredSkill[]`<br>`findByCapability(capability): RegisteredSkill[]` | `SkillProvider`, `FileSkillProvider`, `CloudSkillProvider`, `CompositeSkillProvider` | SkillRegistryV2 component |
-| 70 | `SkillScoring.ts` | `SkillScoring` | `()` | `record(skillId, status)`<br>`getAll(): Record<string, SkillScore>` | — | Tracks skill scores |
-| 71 | `SquadAutoCreator.ts` | `SquadAutoCreator` | `(private baseDir: string = process.env.OPENLIFE_CLOUD_ASSETS_DIR \|\| '.artifacts/cloud-assets')` | `create(goal, squadId): SquadAssetBundle` | — | Auto-creates squad |
-| 72 | `SquadRegistry.ts` | `SquadRegistry` | `(private provider: SquadProvider = new CompositeSquadProvider( new CloudSquadProvider(process.env....)` | `all(): RegisteredSquad[]`<br>`findByCapability(capability): RegisteredSquad[]` | `SquadProvider`, `FileSquadProvider`, `CloudSquadProvider`, `CompositeSquadProvider` | Asset registry for squad |
-| 73 | `SquadRouter.ts` | `SquadRouter` | `(private registry: SquadRegistry)` | `route(task, input): SquadRoute` | `ClassifiedTask`, `TaskIntent`, `SquadRegistry`, `RegisteredSquad` | Routes squad requests to handlers |
-| 74 | `SquadScoring.ts` | `SquadScoring` | `()` | `record(squadId, status)`<br>`getAll(): Record<string, SquadScore>` | — | Tracks squad scores |
-| 75 | `SubagentLifecycle.ts` | `SubagentLifecycle` | `(private filePath: string = process.env.OPENLIFE_CLOUD_SUBAGENTS_FILE \|\| '.artifacts/cloud-subagen...)` | `create(name, capabilities): SubagentRecord`<br>`transition(id, state): SubagentRecord \| null` | — | Subagent lifecycle tracker |
-| 76 | `SynthesizerAgent.ts` | `SynthesizerAgent` | `(private brain: Brain)` | `synthesize(state, lastOutput, review): Promise<string>` | `Brain`, `ExecutionState`, `ReviewFinding` | Synthesizer agent (LLM-backed) |
-| 77 | `SystemDoctor.ts` | `SystemDoctor` | `(private root: string = process.cwd())` | `run(): DoctorCheck[]`<br>`runRuntimeChecks(): Promise<DoctorCheck[]>` | `RuntimeProbe` | Diagnoses System health |
-| 78 | `TaskExecutor.ts` | `TaskExecutor` | `()` | `execute(kind, prompt, projectName): Promise<ExecutionResult>`<br>`executeWithCodex(prompt, projectName): Promise<ExecutionResult>`<br>`executeWithGemini(prompt, projectName): Promise<ExecutionResult>`<br>`executeWithClaude(prompt, projectName, cliName): Promise<ExecutionResult>`<br>`maskSensitiveText(text): string` | — | Executes task |
-| 79 | `TeammateBoard.ts` | `TeammateBoardStore` | `(private workspaceId = 'default')` | `upsert(card): void` | — | Persistent store for teammateboard |
-| 80 | `TestHarness.ts` | `TestHarness` | `()` | `runPhase1Checks(): Promise<HarnessResult[]>` | `Brain`, `Gatekeeper`, `VoiceManager`, `ModelManager`, `Gateway` | Test test harness |
-| 81 | `VoiceManager.ts` | `VoiceManager` | `()` | `processAudioInput(fileLink): Promise<string>`<br>`generateSpeech(text, voiceId): Promise<string>`<br>`logMediaRoute(kind, provider, status, detail): void` | `VoiceRouter` | Manages voice |
-| 82 | `VoiceRouter.ts` | `VoiceRouter` | `()` | `getConfig(): VoiceConfig`<br>`getProfile(profileName): VoiceProfile`<br>`getTtsOrder(profileName): string[]`<br>`getSttOrder(profileName): string[]`<br>`getDefaultVoice(profileName): string \| undefined` | — | Routes voice requests to handlers |
-
-## `providers/` subdirectory
-
-Pluggable asset providers, one class per entity × source. The `Composite*Provider` chains the primary provider over a fallback (typically `File*` over `Cloud*`).
-
-| Entity | File provider | Cloud provider | Composite (chains primary → fallback) |
-|--------|---------------|----------------|----------------------------------------|
-| Agent  | `FileAgentProvider` | `CloudAgentProvider` | `CompositeAgentProvider` |
-| Skill  | `FileSkillProvider` | `CloudSkillProvider` | `CompositeSkillProvider` |
-| Squad  | `FileSquadProvider` | `CloudSquadProvider` | `CompositeSquadProvider` |
-
-**Common surface (all 9 provider classes):**
-- Constructor varies by class:
-  - `File*Provider(roots: string[])` — reads from local `.catalog/` roots
-  - `Cloud*Provider(cloudFile: string)` — reads from `OPENLIFE_CLOUD_*_FILE` env var
-  - `Composite*Provider(primary, fallback)` — chains two providers
-- Public method: `load(): {Agent|Skill|Squad}Record[]`
-
-Type-only files in providers/: `AgentProvider.ts`, `SkillProvider.ts`, `SquadProvider.ts` (interface definitions).
-
-## Summary
-
-- Total files in `src/orchestrator/`: **79**
-- Total files in `src/orchestrator/providers/`: **12**
-- Classes cataloged in Core Classes table: **82**
-- Classes cataloged in `providers/` table (aggregate row): **9**
-- **Total classes:** **91**
-- Files with multiple classes: `EnterpriseAgenticCore.ts` (6 classes — InstitutionalContextStore, PluginManifestStore, PilotStore, MissionEvaluationStore, LearningLoopStore, EnterpriseDoctor)
-- Type-only / interface files (no `export class`): `ExecutionIntent.ts`, `ExecutionState.ts`, `providers/AgentProvider.ts`, `providers/SkillProvider.ts`, `providers/SquadProvider.ts` (5 total)
-
-## Anomalies
-
-- **Type-only files** (export only types/interfaces; no class):
-  - `ExecutionIntent.ts` — exports `ExecutionIntent` type alias and helpers
-  - `ExecutionState.ts` — exports `ExecutionState` type and `createExecutionState` factory function
-  - `providers/AgentProvider.ts` — interface `AgentProvider` only
-  - `providers/SkillProvider.ts` — interface `SkillProvider` only
-  - `providers/SquadProvider.ts` — interface `SquadProvider` only
-- **Classes with no public methods beyond the constructor:** none detected — every class exposes at least one public method.
-- **Naming irregularity:** several stores use `*Store` suffix on the class while the file is named without it (e.g. `JobLifecycle.ts` exports `JobLifecycleStore`, `MissionState.ts` exports `MissionStateStore`, `ServiceState.ts` exports `ServiceStateStore`, `SkillLearningLoop.ts` exports `SkillLearningLoopStore`, `TeammateBoard.ts` exports `TeammateBoardStore`). Also `AgentTeam.ts` exports `AgentTeamRegistry`, `SkillNetwork.ts` exports `SkillNetworkRegistry`. These differ from the file basename so tests should import by class name, not by inferring from filename.
-- **Multi-class file:** `EnterpriseAgenticCore.ts` exports 6 sibling stores under the Enterprise Agentic Core umbrella. Each is independent and should be tested individually.
-- **Truncated constructor signatures** in the table (suffix `...`) indicate default parameter expressions longer than 100 chars (typically `new CompositeXProvider(new CloudXProvider(process.env.OPENLIFE_CLOUD_..._FILE || '...'), new FileXProvider([...]))`). The full default chains the cloud provider over a file provider — see `AgentRegistry.ts`, `SkillRegistryV2.ts`, `SquadRegistry.ts` source for verbatim.