open-multi-agent-kit 0.96.1 → 0.97.0

package/docs/compaction.md

@@ -95,7 +95,7 @@ By default, compaction uses the active session model. Set `compaction.model` to
 
 Every compaction and branch-summary LLM call flows through one choke point (`completeSummarization`):
 
-- **Retries**: transient stream drops (`terminated`, socket close, 5xx, DNS/transport errors) follow the configured `retry` settings (`enabled`, `maxRetries`, `baseDelayMs`) with exponential backoff instead of failing the whole operation on the first attempt. Quota/billing errors fail fast, and aborts are never retried. Retry progress is emitted as `summarization_retry_scheduled` / `summarization_retry_attempt_start` / `summarization_retry_finished` session events (surfaced in the TUI and RPC stream).
+- **Retries**: transient stream drops (`terminated`, socket close, 5xx, DNS/transport errors) follow the configured `retry` settings (`enabled`, `maxRetries`, `baseDelayMs`) with exponential backoff instead of failing the whole operation on the first attempt. Quota/billing exhaustion on the summarization model triggers one pass through the resilience failover chain ([provider-resilience.md](provider-resilience.md)); only when every candidate is also quota-blocked does the run fail with a non-retryable `compaction.quota_exhausted` termination cause whose guidance suggests `/model`, setting `compaction.model`, or waiting for reset. Aborts are never retried. Retry progress is emitted as `summarization_retry_scheduled` / `summarization_retry_attempt_start` / `summarization_retry_finished` session events (surfaced in the TUI and RPC stream).
 - **Isolation**: each summarization request runs with prompt caching disabled (`cacheRetention: "none"`) and a fresh routing `sessionId`, so summaries never write unusable provider cache entries or inherit interactive session affinity.
 
 ### How It Works

package/docs/extensions.md

@@ -505,15 +505,24 @@ The `systemPromptOptions` field gives extensions access to the same structured d
 
 Inside `before_agent_start`, `event.systemPrompt` and `ctx.getSystemPrompt()` both reflect the chained system prompt as of the current handler. Later `before_agent_start` handlers can still modify it again.
 
-#### agent_start / agent_end
+#### agent_start / agent_end / agent_settled
 
-Fired once per user prompt.
+`agent_start` and `agent_end` fire once per agent-loop attempt. When the session
+retries a transient provider failure, the prompt produces several `agent_end`
+events, so `agent_end` alone does not mark the end of the work.
+
+`agent_settled` fires once, immediately after the `agent_end` that no retry
+follows. Use it to release per-run state exactly once.
 
 ```typescript
 omk.on("agent_start", async (_event, ctx) => {});
 
 omk.on("agent_end", async (event, ctx) => {
-  // event.messages - messages from this prompt
+  // event.messages - messages from this attempt; a retry may still follow
+});
+
+omk.on("agent_settled", async (event, ctx) => {
+  // No retry follows. Safe to clear activity, finalize queues, or release locks.
 });
 ```
 

package/docs/index.md

@@ -1,6 +1,8 @@
 # OMK Documentation
 
-OMK is a provider-neutral coding agent with multi-agent orchestration, durable harness controls, replayable sessions, SDK/RPC integration, and evidence-backed verification. Its core stays small while extensions, skills, prompts, themes, and packages add workflows.
+OMK is a provider-neutral coding agent with multi-agent orchestration, durable harness controls, replayable sessions, SDK/RPC integration, and evidence-backed verification. Extensions, skills, prompts, themes, and packages add workflows without changing the run contract.
+
+OMK targets state-of-the-art quality as a CLI coding-agent harness. That target is not a claim of current leadership; see [Turn metrics and harness evaluation](metrics.md) for the qualification rules and current status.
 
 ## Quick start
 
@@ -59,6 +61,7 @@ For the full first-run flow, see [Quickstart](quickstart.md).
 
 - [SDK and Session Control](sdk.md) - embed omk, reuse policy helpers, or inspect stored sessions from scripts.
 - [Run Protocol and Durable Goals](run-protocol.md) - canonical run contracts and the durable-goal lifecycle.
+- [Turn metrics and harness evaluation](metrics.md) - cost, latency, tool reliability, capability baselines, and comparative-claim rules.
 - [RPC mode](rpc.md) - integrate over stdin/stdout JSONL.
 - [JSON event stream mode](json.md) - print mode with structured events.
 - [TUI components](tui.md) - build custom terminal UI for extensions.

package/docs/metrics.md

@@ -3,6 +3,76 @@
 OMK records one JSON line per agent turn so harness changes can be measured
 instead of guessed.
 
+## CLI harness SOTA target
+
+OMK targets state-of-the-art quality as a CLI coding-agent harness. **SOTA is
+not verified.** This is a product target, not a statement that the current
+release leads a benchmark or a named competitor cohort.
+
+The target covers the harness layer:
+
+| Dimension | Primary measure |
+| --- | --- |
+| Task success | solved tasks and pass rate |
+| Cost efficiency | model cost and tokens per solved task |
+| Latency | wall-clock p50/p95 per solved task |
+| Context efficiency | input, cache, compaction, and tool-output tokens per solved task |
+| Tool reliability | failure, retry, refusal, and intervention rates |
+| Orchestration | critical-path time, useful concurrency, and duplicate work |
+| Recovery | interrupted-run resume accuracy and repeated-run variance |
+| Safety | policy violations, unauthorized effects, and false-positive blocks |
+| Maintainability | complexity, module-size debt, and regression-gate health |
+
+No single feature count or self-score establishes leadership. OMK reports a
+dimension-specific result unless a preregistered aggregation defines an overall
+score.
+
+### Controlled comparison contract
+
+A comparative harness run MUST hold the **same model**, **same provider** and
+model configuration, **same task** and revision, **same budget**, equivalent
+tool permissions, and comparable container, hardware, region, and concurrency
+constant. The harness is the treatment variable. If a factor cannot be held
+constant, the report must label the result non-comparative.
+
+Every comparative report must include:
+
+- the date, harness versions, and **named comparison cohort**, with its inclusion rule frozen before execution;
+- immutable model, provider, container, environment, and configuration identities;
+- the task manifest, seeds, public prompt or sanitized prompt digest, budgets, run order, and stop policy;
+- sanitized per-task outcomes plus cost, token, latency, retry, and intervention data;
+- the confidence interval, significance threshold, minimum effect, and statistical test chosen before inspecting the result;
+- the exact commands and immutable manifests needed for **reproducible evidence**.
+
+Use paired A/B measurements. Randomize or interleave pair order when provider or
+machine drift can bias one side. Do not combine values produced by different
+methods, working directories, task revisions, or warm/cold conditions. A
+roadmap projection remains a hypothesis even when its inputs are measured.
+
+### Evidence privacy
+
+Benchmark evidence is private by default. Public artifacts may contain public
+or synthetic task identifiers, version and image digests, allowlisted runtime
+metadata, sanitized outcomes, and aggregate statistics. They must not contain
+credentials, private prompts, proprietary source, raw tool arguments or output,
+environment values, personal data, or absolute user paths.
+
+Keep restricted raw evidence local or in an access-controlled store with an
+explicit owner and retention period. Normalize paths, redact content, scan for
+secrets and PII, and obtain human approval before publication. Digests and
+public task manifests preserve reproducibility without disclosing restricted
+content.
+
+A public “SOTA,” “best,” “leading,” or “#1” claim requires a dated controlled
+comparison that places OMK on the relevant quality/cost/latency frontier without
+violating declared safety and regression floors. Until then, use “targets
+state-of-the-art quality.”
+
+[Harbor's Terminal-Bench runner](https://www.harborframework.com/docs/tutorials/running-terminal-bench)
+and the [SWE-bench containerized harness](https://www.swebench.com/SWE-bench/api/harness/)
+are reference evaluation surfaces. Their presence or a selected task list is
+infrastructure, not a benchmark result.
+
 This is separate from the two things that already existed:
 
 | Surface | Purpose |
@@ -83,5 +153,7 @@ node scripts/tb-mini-suite.mjs --seed 7    # a different fixed subset
 
 Selection is a pure function of (tasks directory, seed, size): the same inputs
 always produce the same task list, which is the whole point of using it as a
-regression gate. The scoring run itself requires Docker, `harbor`, and real
-model spend — it is deliberately not wired into `npm run check`.
+regression gate. Selection alone is not a capability result. The scoring run
+itself requires Docker, `harbor`, and real model spend — it is deliberately not
+wired into `npm run check`. Any comparison produced from it must follow the
+controlled comparison contract above.

package/docs/provider-resilience.md

@@ -6,6 +6,7 @@ OMK can recover an agent turn from provider failures that are unlikely to succee
 - billing-cycle or quota exhaustion
 - orphaned `tool_call_id` protocol errors
 - transient transport and server failures
+- gateway/upstream availability failures (5xx passthroughs, streams that end without a finish reason)
 
 This is availability behavior, not a safety bypass. Provider safety policy and the user's configured model access remain authoritative.
 
@@ -53,6 +54,8 @@ A content/safety stop gets at most one automatic retry, including a retry that s
 
 Recognized quota shapes include billing-cycle usage limits, `insufficient_quota`, exhausted balances, `GoUsageLimitError`, `FreeUsageLimitError`, and out-of-budget responses. These are classified as `provider.rate_limit`, even when a provider wraps them in HTTP 403.
 
+Gateway/upstream availability failures — "503 Upstream request failed", "Endpoint is unavailable", or a stream that ended without a finish reason — are classified as `provider.network`: transport problems that heal by retry or model switch, never by transcript sanitization.
+
 The default candidate order is:
 
 1. `kimi-coding/k3`
@@ -63,6 +66,14 @@ The default candidate order is:
 6. `modelstudio-maas/deepseek-v4-pro`
 7. `kimi-coding/kimi-for-coding`
 
+## Same-model route rotation
+
+The same underlying model is often served by several provider routes (for example `openrouter/stealth/ox-alpha` and `opencode-go/ox-alpha-free`). When the active route fails with an upstream availability error, OMK rotates to another authenticated route of the same model family before touching the cross-model failover chain, so the retry lands on a live endpoint instead of hammering the dead one.
+
+Routes already visited this turn and routes without configured authentication are excluded from rotation. Route rotation runs inside the normal retry budget; when no sibling route remains, the standard failover behavior applies.
+
+Compaction summarization reuses the same chain: when the summarization model hits quota/billing exhaustion, the configured failover candidates are tried once each before the run fails with a non-retryable `compaction.quota_exhausted` termination cause. See [Compaction](compaction.md).
+
 ## Retry and termination events
 
 Each provider attempt is journaled separately and emits `session_termination`. A retryable failure is attempt-level when an `auto_retry_start` event follows it. A recovered retry later emits a `completed` termination; an exhausted retry budget leaves the last provider failure as the final termination.

package/docs/providers.md

@@ -28,7 +28,7 @@ When the status sidebar is pinned, its **USAGE** section lists every configured
 
 Codex streaming passively merges `x-codex-primary-*`, `x-codex-secondary-*`, and `codex.rate_limits` signals through the non-blocking `StreamOptions.onRateLimit` observer. These signals supplement missing polling windows only when the Codex service returns them; OMK does not infer a missing 5-hour value from a 7-day value.
 
-Alibaba Model Studio Token Plan is recognized as **QWEN TOKEN PLAN** but shows `console-only quota`. Its [official usage page](https://modelstudio.console.alibabacloud.com/ap-southeast-1?tab=plan&commonbuy=1&orderType=buy#/efm/subscription/token-plan) obtains `per5HourPercentage`, `per5HourResetTime`, `per1WeekPercentage`, and `per1WeekResetTime` through an authenticated Alibaba Cloud console gateway. The plan-specific `sk-sp-*` key does not authorize that console endpoint, and compatible-mode model responses expose token counts but no quota headers. OMK therefore does not copy browser cookies or estimate quota from token counts. Qwen OAuth remains explicit `quota API unavailable`.
+Alibaba Model Studio Token Plan is recognized as **QWEN TOKEN PLAN** and reads its 7-day window through the official [QwenCloud management CLI](https://docs.qwencloud.com/api-reference/preparation/cli): when `qwencloud` is installed and logged in (`npm i -g @qwencloud/qwencloud-cli && qwencloud auth login`), OMK runs `qwencloud usage summary --format json` (override the binary with `QWENCLOUD_CLI`) and shows `token_plan.usedPct` plus the reset countdown. The CLI holds its own OAuth management credential; OMK never sends the plan's `sk-*` inference key anywhere for quota. Without the CLI the entry shows a `connect:` hint instead — the [console usage page](https://modelstudio.console.alibabacloud.com/ap-southeast-1?tab=plan#/efm/subscription/token-plan/personal) requires an Alibaba Cloud console session that OMK deliberately does not scrape, and OMK still never copies browser cookies or estimates quota from token counts. Qwen OAuth remains explicit `quota API unavailable`.
 
 With a stored native `xai` OAuth credential, OMK reads `GET https://cli-chat-proxy.grok.com/v1/billing?format=credits` and shows the weekly SuperGrok pool from `config.creditUsagePercent` plus its reset from `config.currentPeriod.end`. `XAI_API_KEY` is a separate API-billing credential and does not authorize this subscription endpoint.
 

package/docs/sdk.md

@@ -255,10 +255,14 @@ interface PromptOptions {
   images?: ImageContent[];
   streamingBehavior?: "steer" | "followUp";
   source?: InputSource;
+  activeSkillNames?: readonly string[];
+  activeSkillSource?: string;
   preflightResult?: (success: boolean) => void;
 }
 ```
 
+`activeSkillNames` marks additional discovered skills active for this turn; `activeSkillSource` labels their provenance. They merge with global `defaultActiveSkills`, prioritize matching inventory entries, and do not expand authorization or inline full skill instructions.
+
 `preflightResult` is called once per `prompt()` invocation:
 
 - `true` when the prompt was accepted, queued, or handled immediately

package/docs/settings.md

@@ -180,6 +180,67 @@ Built-in defaults are 30 seconds for `read`, `grep`, `find`, and `ls`; 60 second
 
 The DAG preserves source-order result artifacts. `bash`, unknown tools, and extension tools without explicit resource claims remain exclusive. Tool timeout is logical cancellation: OMK closes the tool call and signals cancellation, but arbitrary JavaScript or external side effects may continue if the tool ignores that signal.
 
+### Resource Governor
+
+The governor probes host capacity (memory, workspace disk, V8 heap, system CPU) and evaluates a resource admission decision, surfaced through `/resource [probe|policy]` in the TUI and `omk doctor resources [--json]` headless. In `observe` mode (default) it records decisions without changing behavior; in `adaptive`/`strict` modes each top-level prompt runs a bounded preflight probe and throttles the effective tool concurrency for that run (never above `agent.maxToolConcurrency`), restoring it when the run settles. Heavy-process caps are enforced at the governed bash boundary; the internal subagent-lane launcher is not wired into live child dispatch. Absent settings keep prior behavior unchanged.
+
+| Setting | Type | Default | Description |
+| --------- | ------ | --------- | ------------- |
+| `resourceGovernor.enabled` | boolean | `true` | `false` forces mode `off` |
+| `resourceGovernor.mode` | string | `"observe"` | `off` (disabled), `observe` (record only), `adaptive`/`strict` (throttle per-run tool concurrency under pressure) |
+| `resourceGovernor.maxProbeMs` | number | `300` | Async probe deadline in ms; valid `50..5000` |
+| `resourceGovernor.cpuSampleMs` | number | `180` | CPU two-sample interval in ms; valid `150..250` |
+| `resourceGovernor.constrainedAvailableMemoryMiB` | number | `1536` | Below this effective available memory the host is `constrained` |
+| `resourceGovernor.criticalAvailableMemoryMiB` | number | `512` | Below this it is `critical`; must be `<=` the constrained threshold |
+| `resourceGovernor.constrainedDiskFreeMiB` | number | `4096` | Workspace free-disk threshold for `constrained` |
+| `resourceGovernor.criticalDiskFreeMiB` | number | `1024` | Workspace free-disk threshold for `critical` |
+| `resourceGovernor.constrainedHeapRatio` | number | `0.75` | V8 heap used/limit ratio for `constrained` |
+| `resourceGovernor.criticalHeapRatio` | number | `0.85` | V8 heap ratio for `critical` |
+| `resourceGovernor.busyCpuPercent` | number | `85` | System CPU percent that throttles; CPU alone never reports `critical` |
+| `resourceGovernor.normalMaxToolConcurrency` | number | `4` | Admission tool cap at `normal` pressure |
+| `resourceGovernor.constrainedMaxToolConcurrency` | number | `2` | Admission tool cap at `constrained` pressure |
+| `resourceGovernor.criticalMaxToolConcurrency` | number | `1` | Admission tool cap at `critical` pressure |
+| `resourceGovernor.normalMaxParallelLanes` | number | `4` | Subagent lane cap at `normal` |
+| `resourceGovernor.constrainedMaxParallelLanes` | number | `2` | Subagent lane cap at `constrained` |
+| `resourceGovernor.criticalMaxParallelLanes` | number | `1` | Subagent lane cap at `critical` |
+| `resourceGovernor.normalMaxHeavyProcesses` | number | `2` | Heavy process cap at `normal` |
+| `resourceGovernor.constrainedMaxHeavyProcesses` | number | `1` | Heavy process cap at `constrained` |
+| `resourceGovernor.criticalMaxHeavyProcesses` | number | `1` | Heavy process cap at `critical` |
+
+Validation is explicit and fail-closed: invalid values are reported (in `/resource policy` and `omk doctor resources --json` under `settingsErrors`) and the defaults apply for the failed area. Caps must satisfy `normal >= constrained >= critical` and stay within `1..64`; critical thresholds must not exceed constrained ones. Admission caps never raise user-configured caps (`agent.maxToolConcurrency` keeps precedence; `0` "unlimited" is limited to the admission cap in `adaptive` and `strict` modes).
+
+`OMK_RESOURCE_GOVERNOR=off|observe|adaptive|strict` overrides the mode for one process, e.g. `OMK_RESOURCE_GOVERNOR=off omk` as a rollback switch.
+
+### Completion Sound
+
+Plays a short system sound when a long prompt finally settles (all retries and continuations drained), only in the interactive TUI on a real TTY — never in RPC, JSON, print mode, or CI. Default off. Backends use fixed executables with fixed arguments (macOS `afplay`, Windows/WSL PowerShell system sound, Linux `canberra-gtk-play`/`paplay`/`aplay`), falling back to the terminal bell. Sound failures are diagnostics only and never affect the prompt outcome.
+
+| Setting | Type | Default | Description |
+| --------- | ------ | --------- | ------------- |
+| `notifications.completionSound.enabled` | boolean | `false` | Master switch; env `OMK_COMPLETION_SOUND` set to `0` or `1` overrides per process |
+| `notifications.completionSound.minDurationMs` | number | `5000` | Only chime for prompts that ran at least this long |
+| `notifications.completionSound.onSuccess` | boolean | `true` | Chime on completed prompts |
+| `notifications.completionSound.onFailure` | boolean | `true` | Chime on failed prompts (user aborts never chime) |
+| `notifications.completionSound.terminalBellFallback` | boolean | `true` | Fall back to the terminal BEL when no sound backend works |
+
+```json
+{
+  "notifications": {
+    "completionSound": { "enabled": true, "minDurationMs": 10000 }
+  }
+}
+```
+
+```json
+{
+  "resourceGovernor": {
+    "mode": "observe",
+    "constrainedAvailableMemoryMiB": 2048,
+    "criticalAvailableMemoryMiB": 512
+  }
+}
+```
+
 ### Branch Summary
 
 | Setting | Type | Default | Description |
@@ -296,11 +357,12 @@ Paths in `~/.omk/agent/settings.json` resolve relative to `~/.omk/agent`. Paths
 | `packages` | array | `[]` | npm/git packages to load resources from |
 | `extensions` | string[] | `[]` | Local extension file paths or directories |
 | `skills` | string[] | `[]` | Local skill file paths or directories |
+| `defaultActiveSkills` | string[] | `[]` | Global-only list of up to 64 exact user-scoped skill names marked active in every prompt; full instructions remain on-demand |
 | `prompts` | string[] | `[]` | Local prompt template paths or directories |
 | `themes` | string[] | `[]` | Local theme file paths or directories |
 | `enableSkillCommands` | boolean | `true` | Register skills as `/skill:name` commands |
 
-Arrays support glob patterns and exclusions. Use `!pattern` to exclude. Use `+path` to force-include an exact path and `-path` to force-exclude an exact path.
+Resource path arrays (`extensions`, `skills`, `prompts`, and `themes`) support glob patterns and exclusions. Use `!pattern` to exclude, `+path` to force-include an exact path, and `-path` to force-exclude one. `defaultActiveSkills` accepts exact names only.
 
 #### packages
 
@@ -356,7 +418,7 @@ See [packages.md](packages.md) for package management details.
 
 ## Project Overrides
 
-Project settings (`.omk/settings.json`) override global settings. Nested objects are merged. `contextBudget` is the exception: it is read from global settings only.
+Project settings (`.omk/settings.json`) override global settings. Nested objects are merged. Settings marked global-only, including `contextBudget` and `defaultActiveSkills`, are read from global settings only.
 
 ```json
 // ~/.omk/agent/settings.json (global)

package/docs/skills.md

@@ -77,6 +77,22 @@ For project-level Claude Code skills, add to `.omk/settings.json`:
 
 This is progressive disclosure: only descriptions are always in context, full instructions load on-demand.
 
+### Default active skills
+
+Set `defaultActiveSkills` in global settings to mark selected discovered skills active in every prompt without repeating `!skill:` invocations:
+
+```json
+{
+  "defaultActiveSkills": ["programming", "review-work", "security-review"]
+}
+```
+
+The global-only setting accepts up to 64 exact skill names, not paths or patterns. Only user-scoped discovered skills can satisfy it; a project skill with the same name is not promoted. Project settings cannot replace or clear the list.
+
+OMK adds a compact active-skills marker containing names and, for explicit-only skills hidden from the normal inventory, the location needed to read them. It does not concatenate every `SKILL.md`; descriptions remain in the normal inventory and full instructions stay on-demand. With context budgeting enabled, active inventory entries receive hard priority. `!skill:` and SDK `activeSkillNames` selections merge for that turn; `/skill:` continues to expand the selected skill directly. Unknown or unavailable names have no prompt effect.
+
+This file is an explicit operator choice, so it may activate a skill whose frontmatter otherwise requires direct invocation. Active skills do not expand authorization and apply only when relevant. Keep the list focused because overlapping language or workflow instructions can conflict.
+
 ## Skill Commands
 
 Skills register as `/skill:name` commands and interactive `!` launcher entries:

package/docs/usage.md

@@ -22,7 +22,7 @@ The editor can be replaced temporarily by built-in UI such as `/settings` or by
 | File reference | Type `@` to fuzzy-search project files |
 | Path completion | Press Tab to complete paths |
 | Multi-line input | Shift+Enter, or Ctrl+Enter on Windows Terminal |
-| Images | Paste with Ctrl+V, Alt+V on Windows, or drag into the terminal |
+| Images | Paste with Ctrl+V (Alt+V on Windows) or drag into the terminal; pastes attach as preview chips above the editor and send with your next prompt |
 | Skill/bash launcher | `!` opens skill completion; `!skill:name prompt` invokes a skill; `!omk <role-or-request>` routes through OMK role hubs |
 | Shell command | `! command` runs and sends output to the model |
 | Hidden shell command | `!! command` runs without sending output to the model |

package/examples/extensions/custom-provider-anthropic/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "omk-extension-custom-provider-anthropic",
-  "version": "0.96.1",
+  "version": "0.97.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "omk-extension-custom-provider-anthropic",
-      "version": "0.96.1",
+      "version": "0.97.0",
       "dependencies": {
         "@anthropic-ai/sdk": "0.52.0"
       }

package/examples/extensions/custom-provider-anthropic/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "omk-extension-custom-provider-anthropic",
 	"private": true,
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/custom-provider-gitlab-duo/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "omk-extension-custom-provider-gitlab-duo",
 	"private": true,
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/gondolin/package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "omk-extension-gondolin",
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "omk-extension-gondolin",
-			"version": "0.96.1",
+			"version": "0.97.0",
 			"dependencies": {
 				"@earendil-works/gondolin": "0.12.0"
 			}

package/examples/extensions/gondolin/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "omk-extension-gondolin",
 	"private": true,
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/sandbox/package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "omk-extension-sandbox",
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "omk-extension-sandbox",
-			"version": "0.96.1",
+			"version": "0.97.0",
 			"dependencies": {
 				"@anthropic-ai/sandbox-runtime": "0.0.26"
 			}

package/examples/extensions/sandbox/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "omk-extension-sandbox",
 	"private": true,
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/subagent/agent-capability-router.test.ts

@@ -9,6 +9,7 @@
 import { describe, expect, it } from "vitest";
 import { auditCapabilities, classifyAgent, deriveCapabilities } from "./agent-capability-router.ts";
 import { buildCapabilityCatalog } from "./capabilities.ts";
+import { DOMAIN_PROFILES } from "./domain-profiles.ts";
 
 // Built once at module scope from the live skill tree under the agent
 // directory. This is populated during local verification but empty in CI (the
@@ -18,6 +19,12 @@ import { buildCapabilityCatalog } from "./capabilities.ts";
 const catalog = buildCapabilityCatalog({ agentDir: `${process.env.HOME ?? "/home/yu"}/.omk/agent` });
 const hasSkillCatalog = catalog.skills.size > 0;
 
+describe("domain profile safety", () => {
+	it("does not route agents to offensive jailbreak skills", () => {
+		expect(DOMAIN_PROFILES.flatMap((profile) => profile.skillPool)).not.toContain("offensive-jailbreak");
+	});
+});
+
 describe("classifyAgent", () => {
 	it("classifies by name token: seo-specialist → marketing-content", () => {
 		expect(classifyAgent("seo-specialist", "x")?.id).toBe("marketing-content");

package/examples/extensions/subagent/domain-profiles.ts

@@ -132,7 +132,6 @@ export const DOMAIN_PROFILES: readonly DomainProfile[] = [
 		],
 		skillPool: [
 			"omk-godmod",
-			"offensive-jailbreak",
 			"reverse-skill",
 			"programming",
 			"insane-search",

package/examples/extensions/with-deps/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "omk-extension-with-deps",
-  "version": "0.96.1",
+  "version": "0.97.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "omk-extension-with-deps",
-      "version": "0.96.1",
+      "version": "0.97.0",
       "dependencies": {
         "ms": "2.1.3"
       },

package/examples/extensions/with-deps/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "omk-extension-with-deps",
 	"private": true,
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/npm-shrinkwrap.json

@@ -1,17 +1,17 @@
 {
 	"name": "open-multi-agent-kit",
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "open-multi-agent-kit",
-			"version": "0.96.1",
+			"version": "0.97.0",
 			"license": "MIT",
 			"dependencies": {
-				"omk-agent-core": "^0.96.1",
-				"omk-ai": "^0.96.1",
-				"omk-tui": "^0.96.1",
+				"omk-agent-core": "^0.97.0",
+				"omk-ai": "^0.97.0",
+				"omk-tui": "^0.97.0",
 				"@silvia-odwyer/photon-node": "0.3.4",
 				"canonicalize": "4.0.0",
 				"chalk": "5.6.2",
@@ -27,8 +27,8 @@
 				"typebox": "1.3.11",
 				"undici": "8.10.0",
 				"yaml": "2.9.0",
-				"omk-adaptorch-wpl": "^0.96.1",
-				"omk-protocol": "^0.96.1"
+				"omk-adaptorch-wpl": "^0.97.0",
+				"omk-protocol": "^0.97.0"
 			},
 			"optionalDependencies": {
 				"@mariozechner/clipboard": "0.3.9"
@@ -1285,19 +1285,19 @@
 			}
 		},
 		"node_modules/omk-adaptorch-wpl": {
-			"version": "0.96.1",
-			"resolved": "https://registry.npmjs.org/omk-adaptorch-wpl/-/omk-adaptorch-wpl-0.96.1.tgz",
+			"version": "0.97.0",
+			"resolved": "https://registry.npmjs.org/omk-adaptorch-wpl/-/omk-adaptorch-wpl-0.97.0.tgz",
 			"license": "MIT",
 			"engines": {
 				"node": ">=22.19.0"
 			}
 		},
 		"node_modules/omk-agent-core": {
-			"version": "0.96.1",
-			"resolved": "https://registry.npmjs.org/omk-agent-core/-/omk-agent-core-0.96.1.tgz",
+			"version": "0.97.0",
+			"resolved": "https://registry.npmjs.org/omk-agent-core/-/omk-agent-core-0.97.0.tgz",
 			"license": "MIT",
 			"dependencies": {
-				"omk-ai": "^0.96.1",
+				"omk-ai": "^0.97.0",
 				"ignore": "7.0.6",
 				"typebox": "1.3.11",
 				"yaml": "2.9.0"
@@ -1307,8 +1307,8 @@
 			}
 		},
 		"node_modules/omk-ai": {
-			"version": "0.96.1",
-			"resolved": "https://registry.npmjs.org/omk-ai/-/omk-ai-0.96.1.tgz",
+			"version": "0.97.0",
+			"resolved": "https://registry.npmjs.org/omk-ai/-/omk-ai-0.97.0.tgz",
 			"license": "MIT",
 			"dependencies": {
 				"@anthropic-ai/sdk": "0.91.1",
@@ -1330,16 +1330,16 @@
 			}
 		},
 		"node_modules/omk-protocol": {
-			"version": "0.96.1",
-			"resolved": "https://registry.npmjs.org/omk-protocol/-/omk-protocol-0.96.1.tgz",
+			"version": "0.97.0",
+			"resolved": "https://registry.npmjs.org/omk-protocol/-/omk-protocol-0.97.0.tgz",
 			"license": "MIT",
 			"engines": {
 				"node": ">=22.19.0"
 			}
 		},
 		"node_modules/omk-tui": {
-			"version": "0.96.1",
-			"resolved": "https://registry.npmjs.org/omk-tui/-/omk-tui-0.96.1.tgz",
+			"version": "0.97.0",
+			"resolved": "https://registry.npmjs.org/omk-tui/-/omk-tui-0.97.0.tgz",
 			"license": "MIT",
 			"dependencies": {
 				"get-east-asian-width": "1.6.0",

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "open-multi-agent-kit",
-	"version": "0.96.1",
+	"version": "0.97.0",
 	"description": "OMK (Open Multi-Agent Kit) coding agent CLI with read, bash, edit, write tools and session management",
 	"type": "module",
 	"omkConfig": {
@@ -32,21 +32,19 @@
 	],
 	"scripts": {
 		"clean": "shx rm -rf dist",
-		"build": "node ./scripts/link-workspace-deps.mjs && tsgo -p tsconfig.build.json && shx chmod +x dist/cli.js && npm run copy-assets",
+		"build": "node ./scripts/link-workspace-deps.mjs && tsgo -p tsconfig.build.json && node ../../scripts/prune-dist-orphans.mjs . && shx chmod +x dist/cli.js && npm run copy-assets",
 		"build:binary": "npm --prefix ../tui run build && npm --prefix ../ai run build && npm --prefix ../agent run build && npm run build && bun build --compile ./dist/bun/cli.js ./src/utils/image-resize-worker.ts --outfile dist/omk && npm run copy-binary-assets",
 		"copy-assets": "shx mkdir -p dist/modes/interactive/theme && shx cp src/modes/interactive/theme/*.json dist/modes/interactive/theme/ && shx mkdir -p dist/modes/interactive/assets && shx cp src/modes/interactive/assets/*.png dist/modes/interactive/assets/ && shx mkdir -p dist/core/export-html/vendor && shx cp src/core/export-html/template.html src/core/export-html/template.css src/core/export-html/template.js dist/core/export-html/ && shx cp src/core/export-html/vendor/*.js dist/core/export-html/vendor/",
 		"copy-binary-assets": "shx cp package.json dist/ && shx cp README.md dist/ && shx cp CHANGELOG.md dist/ && shx mkdir -p dist/theme && shx cp src/modes/interactive/theme/*.json dist/theme/ && shx mkdir -p dist/assets && shx cp src/modes/interactive/assets/*.png dist/assets/ && shx mkdir -p dist/export-html/vendor && shx cp src/core/export-html/template.html dist/export-html/ && shx cp src/core/export-html/vendor/*.js dist/export-html/vendor/ && shx cp -r docs dist/ && shx cp -r examples dist/ && shx cp ../../node_modules/@silvia-odwyer/photon-node/photon_rs_bg.wasm dist/",
 		"test": "vitest --run",
 		"shrinkwrap": "node ../../scripts/generate-coding-agent-shrinkwrap.mjs",
 		"prepublishOnly": "npm run clean && npm run build && npm run shrinkwrap",
-		"link-workspace-deps": "node ./scripts/link-workspace-deps.mjs",
-		"prepare": "node ./scripts/link-workspace-deps.mjs",
-		"postinstall": "node ./scripts/link-workspace-deps.mjs"
+		"link-workspace-deps": "node ./scripts/link-workspace-deps.mjs"
 	},
 	"dependencies": {
-		"omk-agent-core": "^0.96.1",
-		"omk-ai": "^0.96.1",
-		"omk-tui": "^0.96.1",
+		"omk-agent-core": "^0.97.0",
+		"omk-ai": "^0.97.0",
+		"omk-tui": "^0.97.0",
 		"@silvia-odwyer/photon-node": "0.3.4",
 		"canonicalize": "4.0.0",
 		"chalk": "5.6.2",
@@ -62,8 +60,8 @@
 		"typebox": "1.3.11",
 		"undici": "8.10.0",
 		"yaml": "2.9.0",
-		"omk-adaptorch-wpl": "^0.96.1",
-		"omk-protocol": "^0.96.1"
+		"omk-adaptorch-wpl": "^0.97.0",
+		"omk-protocol": "^0.97.0"
 	},
 	"overrides": {
 		"rimraf": "6.1.2",