@mono-agent/agent-app 0.20.10 → 0.20.14

package/skills/mono-agent-composer/references/feature-coverage.md

@@ -6,16 +6,16 @@ Every framework capability and how a composed agent reaches it. This table is th
 
 | Capability | Coverage | Where | Registry config ids |
 | --- | --- | --- | --- |
-| Model backends: claude (sdk/cli), codex (cli direct fallback), pi sdk providers (OpenAI, OpenAI-Codex preferred when Pi auth exists, Copilot, OpenRouter, OpenCode-through-Pi, Ollama, LM Studio, ...), plus hand-authored opencode runtime refs (cli, `opencode:<provider>:<model>` via the OpenCode server) | config | `runtime.model` | `runtime.multi-backend` |
+| Model backends: Pi-only `<provider>:<model>` (OpenAI, OpenAI-Codex, Copilot, Anthropic, OpenRouter, OpenCode-through-Pi, Ollama, LM Studio, ...); a legacy `pi:` prefix is canonicalized away | config | `runtime.model` | `runtime.multi-backend` |
 | Subagent delegation via the Agent tool | config | pi runtime only. Requires BOTH `subagents.enabled: true` and `Agent` in `tools.allowedTools`. Each definition needs exactly one of `prompt` or `promptPath`; omitted `allowedTools` means a read-only default set and `"*"` is rejected. Subagents are capped by `maxConcurrent` (5) and `maxPerTurn` (20), never receive Agent/AskUser/channel-send tools, and cannot spawn subagents | `runtime.subagents` |
 | Same-model retries before failover | config | `runtime.retry.primaryAttempts` (default 2) gives the primary a second attempt before the chain advances; per-route `runtime.fallbacks[].attempts` opts a backup in. Only transient provider failures retry — context overflow and bad credentials still advance. Set `primaryAttempts` to 1 to disable | `runtime.retry` |
-| Backup models on retryable provider failure | config | New configs use `runtime.fallbacks[]` with optional per-route effort. Legacy `runtime.fallbackModels` / `MONO_AGENT_FALLBACK_MODELS` remain compatibility inputs with no removal deadline; do not emit them for new agents | `runtime.fallback-models` |
-| Route-safety contract for primary and fallback models | config | `runtime.routeSafety`: `uniform` (default) or `per-route-native` | `runtime.route-safety` |
-| Execution mode (sdk/cli), effort, max turns, workspace | config + cli | `runtime.executionMode`, `runtime.effort` (`none` / `minimal` / `low` / `medium` / `high` / `xhigh` / `max` / `ultra`; `mono-agent init --effort <level>`). Reasoning-capable `pi:*` maps `ultra` to LOW; Pi without reasoning uses OFF. Direct `codex:*` forwards `ultra` unchanged. Mono-agent rejects `ultra` on its Claude SDK route because the pinned SDK public contract ends at `max` (the SDK JavaScript itself forwards the value). The Claude CLI route passes `--effort ultra`, but both tested Claude Code binaries (SDK-bundled 2.1.206 and local 2.1.210) warn that it is unknown, ignore it, and use default effort. Direct OpenCode rejects explicit effort. Ranking above `max` only prevents keyword downgrade. `runtime.maxTurns`, `runtime.workspace` | `runtime.execution-modes`, `runtime.effort`, `runtime.max-turns`, `runtime.workspace` |
-| Tool-permission posture for CLI backends (direct OpenCode asks/rejects unanswered by default; configure explicitly) | config | `runtime.permissionMode` | `runtime.permission-mode` |
+| Backup models on retryable provider failure | config | `runtime.fallbacks[]`, each route owning optional exact effort (omission = provider default) | `runtime.fallback-models` |
+| Effort, max turns, workspace | config + cli | `runtime.effort` (`none` / `minimal` / `low` / `medium` / `high` / `xhigh` / `max` / `ultra`; `mono-agent init --effort <level>`). Reasoning-capable models map `ultra` to LOW; models without reasoning use OFF. `max` degrades to `xhigh` unless the resolved model advertises it. `mono-agent doctor` validates effort against the model's advertised levels and warns, naming the nearest supported level, when a configured value is outside that set. Ranking above `max` only prevents keyword downgrade. `runtime.maxTurns`, `runtime.workspace` | `runtime.effort`, `runtime.max-turns`, `runtime.workspace` |
+| Tool-permission posture | config | `runtime.permissionMode` (validated and forwarded, not consumed by the Pi runtime) | `runtime.permission-mode` |
 | Continuous provider sessions with idle eviction and optional daily rollover | config | `runtime.session.{mode,idleTimeoutMs,rollover,rolloverTimezone,rolloverNotice}` | `runtime.provider-sessions` |
 | Per-channel run admission/execution bounds | config | `concurrency.maxConcurrentRuns`, `concurrency.maxPendingRuns` | `runtime.concurrency` |
-| Local providers (Ollama / LM Studio / OpenAI-compatible) | config | `providers.local[]` | `runtime.local-providers` |
+| Local providers (Ollama / LM Studio / OpenAI-compatible) | config | `providers.<id>.{type,baseUrl,enabled,trustPublicUrl,apiKeyEnv,models,maxAdvertisedModels}`; the legacy `providers.local[]` array is still accepted and migrated on load | `runtime.local-providers` |
+| Provider catalog (declares supported providers, widening model selection to each provider's full catalog) | config | the `providers` map, keyed by provider id — `providers.piAuthPath`, `providers.piNative`, and `providers.local` are reserved keys, never provider ids, and `providers.entries` is the resolved in-memory shape, not a JSON key. `ollama` and `lmstudio` are zero-config autodiscovered via localhost; a route naming any other provider that is not a Pi builtin fails config load until the map declares it with a `baseUrl` | `runtime.providers` |
 | Pi OAuth credentials | config | `providers.piAuthPath` | `runtime.pi-credentials` |
 | Pi-native transport, retry, and durable provider-session tuning | config | `providers.piNative.{transport,piMaxRetries,maxRetryDelayMs,piSessionsRoot}` | `runtime.pi-native-tuning` |
 | Tool-output bloat guard, cost tracking | auto | built into every run | — |
@@ -38,7 +38,7 @@ Every framework capability and how a composed agent reaches it. This table is th
 | Lite memory (FTS keyword recall + rapid-log capture; no external deps) | config | `memory.mode: "lite"`, `path`, `maxBytes`, `writeMode` | `memory.lite` |
 | Semantic embedding provider selection for Journal/BuJo | config + cli | `memory.embeddings.{provider,endpoint,model,dim,apiKeyEnv}`; guided init supports exclusive Ollama or LM Studio discovery and proof | `memory.embeddings-config` |
 | Journal memory (hybrid recall BM25+vector + static canonical salience; needs configured embeddings) | config | `memory.mode: "journal"`, `path`, `memory.embeddings.{provider,endpoint,model,dim,apiKeyEnv}` (`provider: "ollama" \| "lmstudio" \| "openai"`; exclusive, no cross-provider fallback) | `memory.journal` |
-| BuJo memory (journal + LLM capture/reconcile ADD/UPDATE/SUPERSEDE/NOOP + entity graph + auto-scheduled consolidation; needs embeddings + an app-level `memory.llm`) | config | `memory.mode: "bujo"`, `path`; selected Ollama/LM Studio/OpenAI embeddings are independent from explicit `memory.llm` with `provider: "ollama"` (`model`, optional `endpoint`) or `provider: "agent-host"` (`model` is an SDK runtime model ref, optional `executionMode: "sdk"`) — see `docs/memory/index.md` | `memory.bujo` |
+| BuJo memory (journal + LLM capture/reconcile ADD/UPDATE/SUPERSEDE/NOOP + entity graph + auto-scheduled consolidation; needs embeddings + an app-level `memory.llm`) | config | `memory.mode: "bujo"`, `path`; selected Ollama/LM Studio/OpenAI embeddings are independent from explicit `memory.llm` with `provider: "ollama"` (`model`, optional `endpoint`) or `provider: "agent-host"` (`model` is a runtime model ref) — see `docs/memory/index.md` | `memory.bujo` |
 | Supermemory external backend (server-side extraction/consolidation; async ingestion; explicitly installed plugin) | config | `memory.backend: "supermemory"`, `memory.writeMode`, `memory.supermemory.{baseUrl,apiKey,apiKeyEnv,container,timeoutMs,exposeMcpServer}`; install the exact matching `@mono-agent/memory-supermemory` version | `memory.backend-supermemory` |
 | BuJo consolidation auto-scheduler (projection-only `index.md` refresh + empty `future-log.md` stub + duplicate-group reporting; in-app, no external cron needed) | config | `memory.consolidation.{enabled,cron}` (five-field UTC, default `0 */2 * * *`, no hashed `H`); env `MONO_AGENT_MEMORY_CONSOLIDATION_CRON`, `MONO_AGENT_MEMORY_CONSOLIDATION_ENABLED` | `memory.bujo-consolidation` |
 | Memory maintenance CLI | cli | `mono-agent memory <subcommand>` from the agent folder (stats/today/show/search/top/audit/inspect/rebuild/rollback/…). The standalone `memory-bujo <subcommand> <root>` CLI and bin were removed; routine BuJo consolidation runs via the in-app scheduler | — |
@@ -46,6 +46,7 @@ Every framework capability and how a composed agent reaches it. This table is th
 | Memory liveness check (managed tier/provider/model/dimension identity; provider-native typed discovery plus real finite-vector/dimension probe for Ollama or LM Studio; declared auth env; BuJo LLM config + consolidation cadence; no cross-provider fallback) | cli | `mono-agent validate` | — |
 | Memory write modes and per-turn BuJo capture | config | `memory.writeMode`: `disabled`, `append-host-summary`, or `capture`; capture requires `memory.mode: "bujo"` | `memory.write-mode`, `memory.per-turn-capture` |
 | Auto-provisioned read-only `MemoryRecall` tool exposed for every configured memory tier; no chat LLM | config | `config.memory.recallTool.enabled` (`MONO_AGENT_MEMORY_RECALL_TOOL_ENABLED`, default on; explicit false opts out) | `memory.recall-tool` |
+| Agent-callable `Remember` tool that durably stores one explicitly stated fact; deterministic, append-only, no chat LLM; bujo backend and writable stores only; allowlist-gated and rejects credential-bearing text | config | `config.memory.rememberTool.enabled` (`MONO_AGENT_MEMORY_REMEMBER_TOOL_ENABLED`, default on for the bujo backend; explicit false opts out); a restrictive `tools.allowedTools` must name `Remember` | `memory.remember-tool` |
 | In-app memory LLM call timeout | config | `memory.llm.timeoutMs` (`MONO_AGENT_MEMORY_LLM_TIMEOUT_MS`, default 60000) | `memory.llm-timeout` |
 
 ## Tools, MCP, sandbox
@@ -55,13 +56,14 @@ Every framework capability and how a composed agent reaches it. This table is th
 | Allow-all tool policy (omitted / `["*"]` = all tools; `[]` = none) | config | default `tools.allowedTools`; the harness no-policy safety net is `failClosedToolPolicy()` | `tool-policy.allow-all` |
 | Built-in tool allow/deny lists (deny wins, even under allow-all; pi doesn't deny external MCP tools) | config | `tools.allowedTools`, `tools.disallowedTools`; managed built-ins are Read/Write/Edit/Glob/Grep/Exec/Bash/NodeRepl/WebFetch/WebSearch | `tool-policy.allowlist`, `tool-policy.denylist`, `runtime.builtin-tools` |
 | Pi-native Exec/Bash background process jobs with exact-thread wake | config + cli + auto | `processJobs.*` opts in (default off; unsupported on Windows). Only controller-ready exact Slack, Telegram, and web-console turns gain optional `background`; foreground schemas/behavior stay unchanged otherwise. Host-owned queue/runtime/output/depth/retention caps, owner-private records, inherited POSIX process groups, restart interruption, unforgeable-depth normal-tool wakes, Slack/Telegram same-message lifecycle updates, fail-closed live store health, `mono-agent jobs list\|get\|cancel`, and durable web cards are built in. Busy pre-turn wake admission remains pending without consuming its retry budget. JSON-only `unsafeAllowUnprotectedState` is an explicit sandbox-off/all-Pi trusted-host posture: it suppresses ProcessJobs and clear-sessions SRT while retaining registry, lease, private-root, and provider-zero route invariants, and warns that ProcessJobs state plus the operator secret are model-accessible. Commands that daemonize into another group or session are unsupported | `agent-app.process-jobs` |
+| Pi-native `Monitor`/`MonitorStop` streaming watches with per-batch conversation wakes | config + cli | `monitors.*` opts in (default off; additionally requires `processJobs.enabled`, whose protected private-state root, registration proof and origin binding it reuses; unsupported on Windows). A monitor keeps a command alive after the turn returns, treats each stdout line as one event, coalesces within `coalesceMs`, and wakes the exact originating conversation per batch plus once at the end, steering an in-flight run or queuing its own turn and holding no provider slot. Command preparation, workdir rules, environment cleaning, sandbox seam and output redaction are shared with `Bash`, so there is no separate command allowlist. Capacity is counted independently of `processJobs.*`; one wake per monitor is in flight at a time, pending batches are bounded in lines and bytes with oldest-drop accounting reported to the model, and a sustained firehose stops the watch with `rate_limited` plus one terminal wake. A pre-dispatch refusal is re-offered with the same batch and sequence; every other delivery failure is never replayed. Restart marks live monitors `interrupted`, terminates the incarnation-matched group, and owes exactly one recovery wake without re-running a model-authored command. The fenced envelope declares the turn host-raised and its content untrusted, and a `NOTHING_TO_REPORT` reply is suppressed. Availability is Telegram and Slack only — cron, webhook, TUI-direct and the web console never receive the tools. `mono-agent monitors list\|get\|cancel` is the operator surface | `agent-app.monitors` |
 | Local-first public-web research | config + auto | `tools.web.search.{backend,endpoint,codex.model}` selects strict/first-choice loopback SearXNG, structured ChatGPT-subscription Codex search, or public no-credentials fallbacks; `auto` uses that order and rejects irrelevant/domain-mismatched results before stopping. `tools.web.fetch.{render,browserCommand}` keeps static extraction as the default and optionally enables isolated agent-browser rendering; transient-fetch retry behavior is automatic rather than config-bearing | `runtime.web-research` |
 | MCP servers (stdio/sse/http) from a JSON file | config | `tools.mcpConfigPath` | `tool-policy.mcp-servers` |
 | Generated reply files and Pi-native MCP Apps | config + auto | `PublishReplyFile` is included under allow-all or must be named in a restrictive `tools.allowedTools`; it publishes confined files through owner-private integrity storage for native Slack/Telegram upload and web download. MCP Apps appear only when every configured runtime route supports the Pi-owned bridge, and render only in the hardened web console. Files and Apps share a 20-part run cap; retention follows `artifacts.retention.maxAgeDays` | `agent-app.rich-replies` |
 | Durable origin-bound continuations for trusted stdio/loopback-HTTP MCP services | config + auto | `tools.continuationServers` + `continuations.*`; interactive claims pin a bounded immutable origin snapshot before commit, exact rollover buckets are preserved, v3 state is restart-safe, and unavailable/legacy snapshots use a fixed zero-model fallback | `agent-app.durable-continuations` |
 | Adapter-derived send tools for enabled Slack/Telegram adapters | config | auto-available under allow-all once the channel is enabled; a **specific** `tools.allowedTools` must include `SlackSendMessage` / `TelegramSendMessage`; valid `slack.*` / `telegram.*` config and existing adapter allowlists provide credentials and destination bounds; confirmed message posts are idempotently recorded in destination history | `agent-app.adapter-send-tools` |
 | Interaction bridge for adapter-send history, structured blocking asks, and MCP progress | config + auto | `interaction.bridge.{host,port}`, `interaction.askUser.timeoutMs`, `interaction.progress.enabled`; env `MONO_AGENT_INTERACTION_BRIDGE_HOST`, `MONO_AGENT_INTERACTION_BRIDGE_PORT`, `MONO_AGENT_ASK_USER_TIMEOUT_MS`, `MONO_AGENT_PROGRESS_ENABLED`. It auto-starts for configured Slack/Telegram send tools, when `AskUser` is allowed, when an `interaction` block or interaction env override is configured, or when `interaction.progress.enabled` resolves true and `tools.mcpRequestContextServers` names at least one opted project stdio MCP server. AskUser presents all questions together on web and sequential native controls on Slack/Telegram. | `interaction.bridge` |
-| Sandbox on/off + srt engine (Pi-owned tools; direct Codex has its own sandbox, Claude/direct OpenCode reject native mono policy) | config | `sandbox.mode` | `sandbox.mode` |
+| Sandbox on/off + srt engine (Pi-owned tools) | config | `sandbox.mode` | `sandbox.mode` |
 | Network policy (none/localhost/allowlist/all) | config | `sandbox.network.{mode,allowlist}` | `sandbox.network-policy` |
 | Filesystem scopes (readable/writable roots, deny-write globs) | config | `sandbox.readableRoots`, `sandbox.writableRoots`, `sandbox.denyWrite` | `sandbox.filesystem-scopes` |
 | Fallback behavior when srt is unavailable | config | `sandbox.fallback`, `sandbox.unsafeAllowHostProcess` | `sandbox.fallback` |
@@ -76,7 +78,7 @@ Every framework capability and how a composed agent reaches it. This table is th
 | Telegram long polling, chat allowlist, and group trigger boundary | config | `telegram` section; `telegram.groupMode` (`any` or `mention`) and `telegram.stripMentionText` | `telegram.long-polling` |
 | Telegram runtime/command/reaction/button/file interactivity | config + code | built-in per-chat `/model` and `/effort` use configured primary/fallback models (no Telegram config key); `telegram.commands[]`, `telegram.reactions`, `telegram.quietHours`; `AskUser`, non-blocking `TelegramSendMessage.reply_options`, and `TelegramSendFile` | `telegram.interactive` |
 | Telegram inbound audio transcription | config | `telegram.transcription.{endpoint,model,language,timeoutMs}`; opt-in OpenAI-compatible transcription endpoint for voice notes, audio files, and round-video attachments | `telegram.transcription` |
-| Slack (Socket Mode, channel allowlist, mention handling, native runtime controls) | config + code | `slack` section; built-in Block Kit controls use configured primary/fallback choices through DM-wide/thread-local `@agent /model` and `@agent /effort`, plus channel-wide workspace commands `/<bot>-model` and `/<bot>-effort` derived from `auth.test.user` (Slack app registration + `commands` scope; no mono-agent config key) | `slack.socket-mode` |
+| Slack (Socket Mode, channel allowlist, mention handling, native runtime controls) | config + code | `slack` section; optional `slack.unfurlLinks` / `slack.unfurlMedia` control native agent message previews while omission preserves Slack defaults; built-in Block Kit controls use configured primary/fallback choices through DM-wide/thread-local `@agent /model` and `@agent /effort`, plus channel-wide workspace commands `/<bot>-model` and `/<bot>-effort` derived from `auth.test.user` (Slack app registration + `commands` scope; no mono-agent config key) | `slack.socket-mode` |
 | Slack speaker names (who sent each turn) | config | `slack.resolveUserNames` (default `true`) resolves the sender's display name and handle via `users.info`; requires the `users:read` bot scope; best-effort, so a missing scope leaves turns unnamed instead of failing them; the resolved name is durable (stored turn + memory label) | `slack.speaker-names` |
 | Surface awareness (which channel/DM the turn is in) | config | Always on for Slack and Telegram: the Session block states the surface kind, its id, and the per-message character budget. `slack.resolveChannelNames` (default `true`) adds the Slack channel NAME via `conversations.info` and requires `channels:read`/`groups:read`; best-effort, so a missing scope leaves the surface named by kind and id. Surface ids are model-visible by design — pair with an explicit `slack.allowedChannelIds` allowlist if `SlackSendMessage` is enabled | `channels.surface-awareness` |
 | Slack thread/channel turn context (what was said before the agent was triggered) | config | `slack.threadContext.{enabled,maxMessages,requestLimit,timeoutMs,includeBotMessages}`, defaults `true`/`15`/`15`/`4000`/`true`; reads `conversations.replies` in a thread and `conversations.history` otherwise; requires `channels:history` / `groups:history` / `im:history` / `mpim:history`; one request per turn with a per-channel rate-limit breaker; best-effort, so a missing scope or rate limit sends no transcript instead of failing the turn | `slack.thread-context` |

package/skills/mono-agent-composer/references/package-map.md

@@ -22,7 +22,7 @@ Every real host needs these concepts:
 | --- | --- | --- | --- |
 | Shared request/response shape | `@mono-agent/agent-contracts` | `AgentResponder`, request, response, stream, cancellation contracts | Prompt building, runtime execution, transport |
 | Core config | `@mono-agent/config` | Runtime, context, memory, tools, artifact, traceability settings | Adapter credentials, chat allowlists |
-| Runtime facade | `@mono-agent/runtime-adapter` | Model refs, execution-mode validation, local provider runtime options | Prompts, adapters, memory |
+| Runtime facade | `@mono-agent/runtime-adapter` | Canonical `<provider>:<model>` refs, the Pi backend descriptor, sandbox policy, local provider runtime options | Prompts, adapters, memory |
 | Configured responder | `@mono-agent/agent-app` | Turns `MonoAgentConfig` into runtime, memory, harness, and responder | Polling chats, serving APIs, adapter settings |
 
 Minimal local host:
@@ -87,18 +87,19 @@ request-scoped tools, not entries for `mcp.json`. `SessionHistory` searches the
 current logical session's retained managed-tool calls/results and needs no config
 key. `SetConversationTitle` appears only for writable interactive web threads;
 it keeps automatic semantic titles current, while a user rename permanently
-wins. Allow-all exposes eligible tools on compatible routes; a specific
-allowlist must name each one. Direct OpenCode suppresses all three; direct ACP
-retains and cold-projects lifecycle evidence but cannot expose `SessionHistory`.
+wins. Allow-all exposes eligible tools; a specific allowlist must name each one.
+Every route is Pi-native, so no route family suppresses any of the three; the
+surface conditions above (writable interactive web thread, retained managed-tool
+calls) are the only gates.
 `@mono-agent/agent-app` also owns rich reply composition. `PublishReplyFile`
 copies a confined generated file into owner-private integrity storage and is
 available under allow-all (or by exact name in a restrictive policy). Supported
 Pi-native MCP tool results can register their declared MCP App UI resource for
-the web console. An all-Pi route chain is required for Apps; direct or fallback
-routes that cannot carry the bridge do not advertise it. Adapters consume the
-shared reply-part contract: Slack and Telegram confirm native uploads, the web
-serves authorized downloads and sandboxed Apps, and machine/verbatim adapters
-preserve answer text when they cannot represent a part.
+the web console, and because every route is Pi-native the whole fallback chain
+carries that bridge. Adapters consume the shared reply-part contract: Slack and
+Telegram confirm native uploads, the web serves authorized downloads and
+sandboxed Apps, and machine/verbatim adapters preserve answer text when they
+cannot represent a part.
 
 ```ts
 import { createToolPolicy, toolPolicyToRuntimeOptions } from "@mono-agent/agent-harness";
@@ -129,7 +130,7 @@ Communication adapters are edge packages. They accept an `AgentResponder` and ow
 | Webhook | `@mono-agent/webhook-adapter` | `curl` the configured invocation path (with `Authorization: Bearer ...` when `apiKey` is set) |
 | Cron | `@mono-agent/cron-adapter` | One scheduled or manually triggered invocation |
 
-Adapters must not import the harness, runtime adapter, memory package (`@mono-agent/memory` subpaths), or other adapters. `@mono-agent/agent-app` composes them from config; custom hosts and demos may compose them directly.
+Adapters must not import the harness, runtime adapter, memory package (`@mono-agent/memory` subpaths), or other adapters. `@mono-agent/agent-app` composes them from config; custom app hosts may compose them directly.
 
 ## Observability Join
 
@@ -150,11 +151,16 @@ Use `@mono-agent/agent-orchestrator` when one runtime should call named collabor
 
 Use `@mono-agent/runtime-adapter` unless the host has a custom runtime implementation that already satisfies `MonoRuntimeLike`.
 
-Supported model reference families:
-
-- `codex:<model>` with CLI execution
-- `claude:<model>` with SDK or CLI execution
-- `pi:<provider>:<model>` with SDK execution, including Pi OpenAI-Codex, OpenCode-through-Pi (`pi:opencode-go:*`), and local providers such as Ollama and LM Studio
-- `opencode:<provider>:<model>` with CLI execution for hand-authored OpenCode backend config
+There is one supported model reference family: canonical `<provider>:<model>`,
+run Pi-native. Only the first colon is structural, so `ollama:gemma4:31b` is
+provider `ollama` and model `gemma4:31b`. `<provider>` is a Pi provider id —
+`openai-codex`, `anthropic`, `openai`, `github-copilot`, `openrouter`,
+`opencode-go`, or a local/self-hosted id such as `ollama` or `lmstudio`.
+
+A leading `pi:` is canonicalized away, so `pi:ollama:gemma4:31b` still loads —
+but emit the canonical form. `codex:`, `claude:`, `claude-code:`, `codex-cli:`,
+`acp:`, `vercel:` and the nested `opencode:<provider>:<model>` form were
+retired in 0.21.0 and are rejected at load with the replacement named. Tier
+aliases (`haiku`, `sonnet`, `opus`) are not model ids; use an exact model id.
 
 Provider credentials belong in the provider/runtime environment, not in committed config.

package/skills/mono-agent-composer/references/playbooks.md

@@ -20,7 +20,7 @@ Put `MONO_AGENT_TELEGRAM_BOT_TOKEN=...` in `.env`; the source config omits the c
 
 ```json
 {
-  "runtime": { "model": "pi:openai-codex:gpt-5.6-terra" },
+  "runtime": { "model": "openai-codex:gpt-5.6-terra" },
   "telegram": { "enabled": true, "allowedChatIds": ["123456789"] },
   "memory": {
     "mode": "bujo", "path": "./.mono-agent/memory", "writeMode": "capture",
@@ -30,7 +30,7 @@ Put `MONO_AGENT_TELEGRAM_BOT_TOKEN=...` in `.env`; the source config omits the c
   }
 }
 ```
-**Steps:** `ollama pull nomic-embed-text:v1.5 && ollama pull qwen3.6:latest` → `mono-agent init --model claude:claude-sonnet-4-6 --memory bujo` → add telegram + fill embeddings/llm + `writeMode: capture` → `mono-agent validate` (confirm memory liveness + consolidation cadence) → `mono-agent start`.
+**Steps:** `ollama pull nomic-embed-text:v1.5 && ollama pull qwen3.6:latest` → `mono-agent init --model anthropic:claude-sonnet-4-6 --memory bujo` → add telegram + fill embeddings/llm + `writeMode: capture` → `mono-agent validate` (confirm memory liveness + consolidation cadence) → `mono-agent start`.
 **Smoke:** send a fact from the allowed chat, then ask a paraphrased question later; confirm the final answer arrives separately and the temporary memory-tool activity disappears, `MemoryRecall` appears in the run JSONL, and the answer uses it.
 
 ## 2. Slack team bot with MCP tools
@@ -42,7 +42,7 @@ Put `MONO_AGENT_SLACK_BOT_TOKEN` and `MONO_AGENT_SLACK_APP_TOKEN` in `.env`; the
 
 ```json
 {
-  "runtime": { "model": "pi:openai-codex:gpt-5.6-terra" },
+  "runtime": { "model": "openai-codex:gpt-5.6-terra" },
   "slack": { "enabled": true, "allowedChannelIds": ["C012345"], "stripMentionText": true },
   "tools": { "allowedTools": ["Read", "Grep", "SlackSendMessage", "deployTool"], "mcpConfigPath": "./mcp.json" },
   "concurrency": { "maxConcurrentRuns": 4, "maxPendingRuns": 8 }
@@ -58,13 +58,13 @@ Put `MONO_AGENT_SLACK_BOT_TOKEN` and `MONO_AGENT_SLACK_APP_TOKEN` in `.env`; the
 
 ```json
 {
-  "runtime": { "model": "pi:ollama:gemma4:31b" },
+  "runtime": { "model": "ollama:gemma4:31b" },
   "providers": { "local": [{ "id": "ollama", "type": "ollama", "baseUrl": "http://localhost:11434", "enabled": true, "models": [{ "name": "gemma4:31b", "capabilities": { "context_window": 32768 } }] }] },
   "memory": { "mode": "journal", "path": "./.mono-agent/memory", "embeddings": { "provider": "ollama", "model": "nomic-embed-text:v1.5", "endpoint": "http://localhost:11434", "dim": 768 } },
   "sandbox": { "mode": "native", "network": { "mode": "localhost" } }
 }
 ```
-**Steps:** pull both models → `mono-agent init --model pi:ollama:gemma4:31b --memory journal` → add `providers.local` + embeddings + `sandbox.network.mode: localhost` → `validate` (Ollama reachable, models pulled) → `start`.
+**Steps:** pull both models → `mono-agent init --model ollama:gemma4:31b --memory journal` → add `providers.local` + embeddings + `sandbox.network.mode: localhost` → `validate` (Ollama reachable, models pulled) → `start`.
 **Smoke:** `curl -X POST` the webhook path; confirm a local-model response and no outbound non-localhost network in the artifact.
 
 ## 4. OpenAI-compatible endpoint for Open WebUI
@@ -74,7 +74,7 @@ Put `MONO_AGENT_SLACK_BOT_TOKEN` and `MONO_AGENT_SLACK_APP_TOKEN` in `.env`; the
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6", "session": { "mode": "continuous", "idleTimeoutMs": 1800000 } },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6", "session": { "mode": "continuous", "idleTimeoutMs": 1800000 } },
   "openaiApi": { "enabled": true, "host": "0.0.0.0", "port": 4040, "basePath": "/v1", "allowNonLoopback": true, "modelId": "my-agent" }
 }
 ```
@@ -88,7 +88,7 @@ Put `MONO_AGENT_SLACK_BOT_TOKEN` and `MONO_AGENT_SLACK_APP_TOKEN` in `.env`; the
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6" },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6" },
   "webhook": {
     "enabled": true, "host": "127.0.0.1", "port": 8080, "defaultMode": "sync",
     "endpoints": [
@@ -113,13 +113,13 @@ Put `MONO_AGENT_SLACK_BOT_TOKEN` and `MONO_AGENT_SLACK_APP_TOKEN` in `.env`; the
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6" },
-  "slack": { "enabled": true, "allowedChannelIds": ["C012345"] },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6" },
+  "slack": { "enabled": true, "allowedChannelIds": ["C012345"], "unfurlLinks": false, "unfurlMedia": false },
   "cron": { "jobs": [{ "id": "morning-digest", "enabled": true, "expression": "0 9 * * *", "timezone": "America/New_York", "prompt": "Build the morning digest. Your final answer is the digest to notify.", "conversationId": "daily-digest", "notify": true, "notifyConversationId": "slack:C012345" }] }
 }
 ```
 **Steps:** `mono-agent init` → add the destination adapter and allowlist → add the cron job (or `cron/morning-digest.md`) with `conversationId`, IANA timezone, `notify: true`, and optional `notifyConversationId` → `validate` → `start`.
-**Smoke:** trigger a one-off tick; confirm the final answer lands verbatim in the allowed destination with no tool call and `conversationId` shares context across ticks. Return `NOTHING_TO_REPORT` to test the silent path.
+**Smoke:** trigger a one-off tick; confirm the final answer lands verbatim in the allowed destination with no tool call, no Slack link/media preview under the example config, and `conversationId` shares context across ticks. Return `NOTHING_TO_REPORT` to test the silent path.
 
 ## 7. A2A provider + consumer pair
 **For:** a platform integrator connecting two agents over A2A.
@@ -180,13 +180,13 @@ const orchestrator = await createConfiguredAgentResponder({
 
 ```json
 {
-  "runtime": { "model": "pi:openai-codex:gpt-5.6-terra" },
+  "runtime": { "model": "openai-codex:gpt-5.6-terra" },
   "tools": { "allowedTools": ["*"] },
   "sandbox": { "mode": "native", "network": { "mode": "localhost" }, "readableRoots": ["."], "writableRoots": ["."], "denyWrite": [".env", ".env.*", ".git/config", ".git/hooks/**"], "fallback": "fail-closed" }
 }
 ```
 **Steps:** `mono-agent init --memory journal` → leave tools at the allow-all default (`["*"]`); the **sandbox**, not an allowlist, is what constrains the code tools → `sandbox.mode native` + `network localhost` + deny-write defaults → keep `fallback: fail-closed` (do NOT set `unsafe-host-process`) → `validate` → `start`.
-**Smoke:** ask it to read a file, use Exec for one argv-safe command, run one Bash pipeline, then use NodeRepl twice to retain a variable and produce `42` (all work); next fetch an external URL or write `.env` (both blocked in the artifact). Keep every primary/fallback/trigger model on Pi; direct Codex, Claude, and direct OpenCode reject this mono-agent sandbox policy.
+**Smoke:** ask it to read a file, use Exec for one argv-safe command, run one Bash pipeline, then use NodeRepl twice to retain a variable and produce `42` (all work); next fetch an external URL or write `.env` (both blocked in the artifact). Every route is Pi-native, so this sandbox policy applies uniformly to the primary, every fallback, and every per-trigger override.
 
 ## 10. Phoenix-observed agent with the TUI
 **For:** an agent builder evaluating runs in a tracing dashboard.
@@ -195,7 +195,7 @@ const orchestrator = await createConfiguredAgentResponder({
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6" },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6" },
   "artifacts": { "dir": ".mono-agent/artifacts" },
   "traceability": { "registryDir": ".mono-agent/trace-sources", "sourceId": "my-agent", "heartbeatMs": 10000 },
   "observability": { "exporters": [{ "type": "phoenix", "endpoint": "http://127.0.0.1:6006/v1/traces", "projectName": "my-project", "includeSensitiveData": false, "contentPatternRedaction": false, "timeoutMs": 5000 }] }
@@ -223,16 +223,12 @@ const orchestrator = await createConfiguredAgentResponder({
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6", "fallbacks": [{ "model": "pi:openai-codex:gpt-5.5" }, { "model": "pi:ollama:gemma4:31b" }], "session": { "mode": "continuous" } },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6", "fallbacks": [{ "model": "openai-codex:gpt-5.6-sol" }, { "model": "ollama:gemma4:31b" }], "session": { "mode": "continuous" } },
   "providers": { "local": [{ "id": "ollama", "type": "ollama", "baseUrl": "http://localhost:11434", "enabled": true }], "piNative": { "transport": "auto", "piMaxRetries": 2, "maxRetryDelayMs": 60000, "piSessionsRoot": ".mono-agent/sessions" } }
 }
 ```
-Legacy `runtime.fallbackModels` and `MONO_AGENT_FALLBACK_MODELS` remain supported
-with no removal deadline, but this playbook intentionally emits the canonical
-per-route form for a new agent.
-**Steps:** `ollama pull gemma4:31b` → `mono-agent init --model claude:claude-sonnet-4-6 --fallback pi:openai-codex:gpt-5.5 --fallback pi:ollama:gemma4:31b` → add `providers.local` + `piNative.piSessionsRoot` → `validate` → `start`.
-**Boundary:** this mixed Pi/Claude chain intentionally omits the native mono-agent sandbox. Keep direct Codex chains all-direct; keep every route on Pi (including `pi:opencode-go:*`, not direct `opencode:*`) when `sandbox.mode` is `native`.
-**Smoke:** force a retryable primary failure; confirm the run result reports failover to the next model (not silent) and the conversation resumes from the transcript tail.
+`runtime.fallbackModels` and `MONO_AGENT_FALLBACK_MODELS` were retired in 0.21.0
+and are rejected at load; emit `runtime.fallbacks[]` only.
 
 ## 13. Personal Telegram assistant with Supermemory
 **For:** a power user trying an external memory layer while keeping the agent local.
@@ -241,7 +237,7 @@ per-route form for a new agent.
 
 ```json
 {
-  "runtime": { "model": "claude:claude-sonnet-4-6" },
+  "runtime": { "model": "anthropic:claude-sonnet-4-6" },
   "telegram": { "enabled": true, "allowedChatIds": ["123456789"] },
   "memory": {
     "backend": "supermemory", "mode": "lite", "path": "./.mono-agent/memory",
@@ -261,13 +257,13 @@ per-route form for a new agent.
 
 ```json
 {
-  "runtime": { "model": "pi:lmstudio:qwen3.6-32b" },
+  "runtime": { "model": "lmstudio:qwen3.6-32b" },
   "providers": { "local": [{ "id": "lmstudio", "type": "lmstudio", "baseUrl": "http://localhost:1234", "enabled": true }] },
   "memory": { "mode": "lite", "path": "./.mono-agent/memory", "writeMode": "append-host-summary" },
   "webhook": { "enabled": true }
 }
 ```
-**Steps:** start LM Studio's local server with the chosen model loaded → `mono-agent init --model pi:lmstudio:qwen3.6-32b --memory lite` (the `pi:lmstudio:*` model auto-adds the LM Studio provider block; there is no LM Studio preset — `local-private` is Ollama-based) → adjust `runtime.model` if the displayed model id differs → `validate` → `start`.
+**Steps:** start LM Studio's local server with the chosen model loaded → `mono-agent init --model lmstudio:qwen3.6-32b --memory lite` (the `lmstudio:*` model auto-adds the LM Studio provider block; there is no LM Studio preset — `local-private` is Ollama-based) → adjust `runtime.model` if the displayed model id differs → `validate` → `start`.
 **Smoke:** `curl` the webhook invoke URL and confirm the response comes from the local LM Studio model.
 
 ## 15. Interactive agent with long jobs and large media
@@ -279,7 +275,7 @@ per-route form for a new agent.
 
 ```json
 {
-  "runtime": { "model": "pi:openai-codex:gpt-5.5", "executionMode": "sdk" },
+  "runtime": { "model": "openai-codex:gpt-5.6-sol" },
   "tools": { "allowedTools": ["Read", "AskUser", "TelegramSendFile", "PublishReplyFile"], "mcpConfigPath": "./.mcp.json", "mcpCallMaxTotalTimeoutMs": 2700000 },
   "interaction": { "bridge": { "host": "127.0.0.1", "port": 4471 }, "askUser": { "timeoutMs": 600000 }, "progress": { "enabled": true } },
   "telegram": { "enabled": true, "allowedChatIds": ["123456789"], "apiRoot": "http://127.0.0.1:8081", "attachments": { "maxBytes": 268435456, "maxUploadBytes": 268435456 } }
@@ -295,7 +291,7 @@ per-route form for a new agent.
 
 ```json
 {
-  "runtime": { "model": "pi:openai-codex:gpt-5.5" },
+  "runtime": { "model": "openai-codex:gpt-5.6-sol" },
   "tools": {
     "allowedTools": ["Read", "Glob", "Grep", "WebSearch", "WebFetch"],
     "web": {
@@ -307,6 +303,6 @@ per-route form for a new agent.
 }
 ```
 
-**Steps:** start the pinned loopback companion from `demos/searxng` → keep strict `searxng` or choose `auto` for ChatGPT-subscription Codex then keyless fallback → keep render `never`, or install `agent-browser >=0.33.1` and opt into `auto` for SPA pages → `validate` → `start`.
+**Steps:** provision an operator-owned loopback SearXNG instance with JSON responses enabled → keep strict `searxng` or choose `auto` for ChatGPT-subscription Codex then keyless fallback → keep render `never`, or install `agent-browser >=0.33.1` and opt into `auto` for SPA pages → `validate` → `start`.
 **Boundary:** local SearXNG is private infrastructure, not an offline index; it contacts public engines and WebFetch contacts result sites. `localhost` permits the companion but blocks public fetches.
 **Smoke:** ask for two query variants and one official-page fetch; require canonical ranked URLs, untrusted-content boundaries, bounded timing metadata without query/URL leakage, and no duplicate network work for an identical call in the run.

package/skills/mono-agent-composer/references/validation.md

@@ -101,14 +101,11 @@ pnpm --filter @mono-agent/<package> run build
 pnpm --filter @mono-agent/<package> run typecheck
 ```
 
-For broad host or demo changes, run the full workspace gate:
+For broad host or repository changes, run the full workspace gate:
 
 ```bash
 pnpm run build
 pnpm test
-pnpm run build:demo
-pnpm run typecheck:demo
-pnpm run test:demo
 ```
 
 ## Smoke Tests By Surface

package/dist/codex-model-catalog.d.ts

@@ -1,15 +0,0 @@
-import type { EffortLevel } from "@mono-agent/config";
-export interface CodexCatalogModel {
-    readonly id: string;
-    readonly displayName: string;
-    readonly supportedEfforts: readonly EffortLevel[];
-    readonly defaultEffort?: EffortLevel;
-    readonly isDefault?: boolean;
-}
-/** Exact non-secret environment used by live Codex model/list discovery. */
-export declare function codexModelDiscoveryEnvironment(persistedEnv?: Readonly<Record<string, string | undefined>>, shellEnv?: Readonly<Record<string, string | undefined>>): Record<string, string | undefined>;
-export declare function normalizeEffort(value: unknown): EffortLevel | undefined;
-export declare function normalizeEfforts(values: readonly unknown[]): EffortLevel[];
-export declare function normalizeCodexCatalog(rows: readonly unknown[]): CodexCatalogModel[];
-export declare function requestCodexModelList(timeoutMs: number, environment: Readonly<Record<string, string | undefined>>, abortSignal?: AbortSignal): Promise<CodexCatalogModel[]>;
-//# sourceMappingURL=codex-model-catalog.d.ts.map

package/dist/codex-model-catalog.d.ts.map

@@ -1 +0,0 @@
-{"version":3,"file":"codex-model-catalog.d.ts","sourceRoot":"","sources":["../src/codex-model-catalog.ts"],"names":[],"mappings":"AAIA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AAItD,MAAM,WAAW,iBAAiB;IAChC,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAC;IACpB,QAAQ,CAAC,WAAW,EAAE,MAAM,CAAC;IAC7B,QAAQ,CAAC,gBAAgB,EAAE,SAAS,WAAW,EAAE,CAAC;IAClD,QAAQ,CAAC,aAAa,CAAC,EAAE,WAAW,CAAC;IACrC,QAAQ,CAAC,SAAS,CAAC,EAAE,OAAO,CAAC;CAC9B;AAgDD,4EAA4E;AAC5E,wBAAgB,8BAA8B,CAC5C,YAAY,GAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,CAAM,EAC/D,QAAQ,GAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,CAAe,GACnE,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,CAOpC;AAED,wBAAgB,eAAe,CAAC,KAAK,EAAE,OAAO,GAAG,WAAW,GAAG,SAAS,CAGvE;AAED,wBAAgB,gBAAgB,CAAC,MAAM,EAAE,SAAS,OAAO,EAAE,GAAG,WAAW,EAAE,CAE1E;AAED,wBAAgB,qBAAqB,CAAC,IAAI,EAAE,SAAS,OAAO,EAAE,GAAG,iBAAiB,EAAE,CAsBnF;AAED,wBAAsB,qBAAqB,CACzC,SAAS,EAAE,MAAM,EACjB,WAAW,EAAE,QAAQ,CAAC,MAAM,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,CAAC,CAAC,EACzD,WAAW,CAAC,EAAE,WAAW,GACxB,OAAO,CAAC,iBAAiB,EAAE,CAAC,CAoH9B"}

package/dist/codex-model-catalog.js

@@ -1,220 +0,0 @@
-import { spawn } from "node:child_process";
-import { homedir } from "node:os";
-import { createInterface } from "node:readline";
-import { credentialNeutralProviderStatusEnvironment } from "./provider-setup.js";
-const SUPPORTED_EFFORTS = new Set([
-    "none",
-    "minimal",
-    "low",
-    "medium",
-    "high",
-    "xhigh",
-    "max",
-    "ultra",
-]);
-function isRecord(value) {
-    return typeof value === "object" && value !== null && !Array.isArray(value);
-}
-function boundedCatalogText(value, limit) {
-    const normalized = typeof value === "string"
-        ? value.replace(/[\u0000-\u001f\u007f]/gu, " ").replace(/\s+/gu, " ").trim()
-        : "";
-    return normalized.length <= limit ? normalized : `${normalized.slice(0, Math.max(0, limit - 1))}…`;
-}
-function safeDiscoveryProcessEnv(source = process.env) {
-    const env = {};
-    for (const key of [
-        "PATH",
-        "LANG",
-        "LC_ALL",
-        "LC_CTYPE",
-        "SHELL",
-        "TMPDIR",
-        "TMP",
-        "TEMP",
-        "SystemRoot",
-        "WINDIR",
-        "ComSpec",
-        "PATHEXT",
-    ]) {
-        const value = source[key];
-        if (typeof value === "string" && value.length > 0)
-            env[key] = value;
-    }
-    return env;
-}
-/** Exact non-secret environment used by live Codex model/list discovery. */
-export function codexModelDiscoveryEnvironment(persistedEnv = {}, shellEnv = process.env) {
-    const neutral = credentialNeutralProviderStatusEnvironment(shellEnv, persistedEnv);
-    return {
-        ...safeDiscoveryProcessEnv(shellEnv),
-        HOME: neutral.HOME ?? homedir(),
-        ...(neutral.CODEX_HOME === undefined ? {} : { CODEX_HOME: neutral.CODEX_HOME }),
-    };
-}
-export function normalizeEffort(value) {
-    const normalized = value === "off" ? "none" : typeof value === "string" ? value : "";
-    return SUPPORTED_EFFORTS.has(normalized) ? normalized : undefined;
-}
-export function normalizeEfforts(values) {
-    return [...new Set(values.map(normalizeEffort).filter((value) => value !== undefined))];
-}
-export function normalizeCodexCatalog(rows) {
-    const result = [];
-    const seen = new Set();
-    for (const raw of rows.slice(0, 1_000)) {
-        if (!isRecord(raw))
-            continue;
-        const id = typeof raw.id === "string" ? raw.id.trim() : "";
-        if (id.length === 0 || id.length > 160 || seen.has(id))
-            continue;
-        const supportedRows = Array.isArray(raw.supportedReasoningEfforts)
-            ? raw.supportedReasoningEfforts.map((entry) => isRecord(entry) ? entry.reasoningEffort : entry)
-            : [];
-        const supportedEfforts = normalizeEfforts(supportedRows);
-        const normalizedDefault = normalizeEffort(raw.defaultReasoningEffort);
-        seen.add(id);
-        result.push({
-            id,
-            displayName: boundedCatalogText(raw.displayName, 160) || id,
-            supportedEfforts,
-            ...(normalizedDefault === undefined ? {} : { defaultEffort: normalizedDefault }),
-            ...(typeof raw.isDefault === "boolean" ? { isDefault: raw.isDefault } : {}),
-        });
-    }
-    return result;
-}
-export async function requestCodexModelList(timeoutMs, environment, abortSignal) {
-    const child = spawn("codex", ["app-server", "--listen", "stdio://", "-c", "project_doc_max_bytes=0"], {
-        stdio: ["pipe", "pipe", "ignore"],
-        env: { ...environment },
-    });
-    const lines = createInterface({ input: child.stdout });
-    return await new Promise((resolveResult, rejectResult) => {
-        let settled = false;
-        let timer;
-        const ignoreStdinError = () => undefined;
-        child.stdin.on("error", ignoreStdinError);
-        const childClosedWithin = async (milliseconds) => {
-            if (child.exitCode !== null || child.signalCode !== null)
-                return true;
-            return await new Promise((resolveClosed) => {
-                const onClose = () => {
-                    clearTimeout(closeTimer);
-                    resolveClosed(true);
-                };
-                const closeTimer = setTimeout(() => {
-                    child.off("close", onClose);
-                    resolveClosed(false);
-                }, milliseconds);
-                closeTimer.unref?.();
-                child.once("close", onClose);
-            });
-        };
-        const stopChild = async () => {
-            lines.removeAllListeners("line");
-            lines.close();
-            try {
-                child.stdin.end();
-            }
-            catch { /* best effort */ }
-            if (await childClosedWithin(25))
-                return;
-            try {
-                child.kill("SIGTERM");
-            }
-            catch { /* best effort */ }
-            if (await childClosedWithin(250))
-                return;
-            try {
-                child.kill("SIGKILL");
-            }
-            catch { /* best effort */ }
-            if (!await childClosedWithin(250)) {
-                child.stdin.destroy();
-                child.stdout.destroy();
-                child.unref();
-            }
-        };
-        let settle;
-        const onChildError = () => void settle(new Error("Codex app-server is unavailable."));
-        const onChildClose = () => void settle(new Error("Codex app-server closed before returning models."));
-        const onAbort = () => {
-            const error = new Error("Codex model catalog discovery was interrupted.");
-            error.name = "AbortError";
-            void settle(error);
-        };
-        settle = async (error, models = []) => {
-            if (settled)
-                return;
-            settled = true;
-            clearTimeout(timer);
-            child.off("error", onChildError);
-            child.off("close", onChildClose);
-            abortSignal?.removeEventListener("abort", onAbort);
-            await stopChild();
-            child.stdin.off("error", ignoreStdinError);
-            if (error === undefined)
-                resolveResult(models);
-            else
-                rejectResult(error);
-        };
-        const writeMessage = (message) => {
-            if (settled || child.stdin.destroyed || child.stdin.writableEnded)
-                return;
-            try {
-                child.stdin.write(`${JSON.stringify(message)}\n`);
-            }
-            catch { /* bounded shutdown handles the process */ }
-        };
-        timer = setTimeout(() => void settle(new Error("Codex model catalog discovery timed out.")), Math.max(250, Math.min(10_000, timeoutMs)));
-        timer.unref?.();
-        child.once("error", onChildError);
-        child.once("close", onChildClose);
-        abortSignal?.addEventListener("abort", onAbort, { once: true });
-        lines.on("line", (line) => {
-            let message;
-            try {
-                message = JSON.parse(line);
-            }
-            catch {
-                return;
-            }
-            if (!isRecord(message))
-                return;
-            if (typeof message.method === "string" && (typeof message.id === "string" || typeof message.id === "number")) {
-                writeMessage({
-                    id: message.id,
-                    error: { code: -32601, message: "Model catalog discovery does not support server requests." },
-                });
-                return;
-            }
-            if (message.id === 1 && isRecord(message.result)) {
-                writeMessage({
-                    id: 2,
-                    method: "model/list",
-                    params: { includeHidden: false, limit: 1_000 },
-                });
-                return;
-            }
-            if (message.id !== 2)
-                return;
-            if (!isRecord(message.result) || !Array.isArray(message.result.data)) {
-                void settle(new Error("Codex app-server returned an invalid model catalog."));
-                return;
-            }
-            void settle(undefined, normalizeCodexCatalog(message.result.data));
-        });
-        writeMessage({
-            id: 1,
-            method: "initialize",
-            params: {
-                clientInfo: { name: "mono-agent", title: "mono-agent", version: "0" },
-                capabilities: { experimentalApi: true },
-            },
-        });
-        if (abortSignal?.aborted === true)
-            onAbort();
-    });
-}
-//# sourceMappingURL=codex-model-catalog.js.map

package/dist/codex-model-catalog.js.map

@@ -1 +0,0 @@
-{"version":3,"file":"codex-model-catalog.js","sourceRoot":"","sources":["../src/codex-model-catalog.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,KAAK,EAAE,MAAM,oBAAoB,CAAC;AAC3C,OAAO,EAAE,OAAO,EAAE,MAAM,SAAS,CAAC;AAClC,OAAO,EAAE,eAAe,EAAE,MAAM,eAAe,CAAC;AAIhD,OAAO,EAAE,0CAA0C,EAAE,MAAM,qBAAqB,CAAC;AAUjF,MAAM,iBAAiB,GAAG,IAAI,GAAG,CAAc;IAC7C,MAAM;IACN,SAAS;IACT,KAAK;IACL,QAAQ;IACR,MAAM;IACN,OAAO;IACP,KAAK;IACL,OAAO;CACR,CAAC,CAAC;AAEH,SAAS,QAAQ,CAAC,KAAc;IAC9B,OAAO,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,KAAK,IAAI,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,KAAK,CAAC,CAAC;AAC9E,CAAC;AAED,SAAS,kBAAkB,CAAC,KAAc,EAAE,KAAa;IACvD,MAAM,UAAU,GAAG,OAAO,KAAK,KAAK,QAAQ;QAC1C,CAAC,CAAC,KAAK,CAAC,OAAO,CAAC,yBAAyB,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE;QAC5E,CAAC,CAAC,EAAE,CAAC;IACP,OAAO,UAAU,CAAC,MAAM,IAAI,KAAK,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,GAAG,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC;AACrG,CAAC;AAED,SAAS,uBAAuB,CAC9B,SAAuD,OAAO,CAAC,GAAG;IAElE,MAAM,GAAG,GAAuC,EAAE,CAAC;IACnD,KAAK,MAAM,GAAG,IAAI;QAChB,MAAM;QACN,MAAM;QACN,QAAQ;QACR,UAAU;QACV,OAAO;QACP,QAAQ;QACR,KAAK;QACL,MAAM;QACN,YAAY;QACZ,QAAQ;QACR,SAAS;QACT,SAAS;KACV,EAAE,CAAC;QACF,MAAM,KAAK,GAAG,MAAM,CAAC,GAAG,CAAC,CAAC;QAC1B,IAAI,OAAO,KAAK,KAAK,QAAQ,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC;YAAE,GAAG,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC;IACtE,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,4EAA4E;AAC5E,MAAM,UAAU,8BAA8B,CAC5C,eAA6D,EAAE,EAC/D,WAAyD,OAAO,CAAC,GAAG;IAEpE,MAAM,OAAO,GAAG,0CAA0C,CAAC,QAAQ,EAAE,YAAY,CAAC,CAAC;IACnF,OAAO;QACL,GAAG,uBAAuB,CAAC,QAAQ,CAAC;QACpC,IAAI,EAAE,OAAO,CAAC,IAAI,IAAI,OAAO,EAAE;QAC/B,GAAG,CAAC,OAAO,CAAC,UAAU,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,OAAO,CAAC,UAAU,EAAE,CAAC;KAChF,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,KAAc;IAC5C,MAAM,UAAU,GAAG,KAAK,KAAK,KAAK,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,OAAO,KAAK,KAAK,QAAQ,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;IACrF,OAAO,iBAAiB,CAAC,GAAG,CAAC,UAAyB,CAAC,CAAC,CAAC,CAAC,UAAyB,CAAC,CAAC,CAAC,SAAS,CAAC;AAClG,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,MAA0B;IACzD,OAAO,CAAC,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,GAAG,CAAC,eAAe,CAAC,CAAC,MAAM,CAAC,CAAC,KAAK,EAAwB,EAAE,CAAC,KAAK,KAAK,SAAS,CAAC,CAAC,CAAC,CAAC;AAChH,CAAC;AAED,MAAM,UAAU,qBAAqB,CAAC,IAAwB;IAC5D,MAAM,MAAM,GAAwB,EAAE,CAAC;IACvC,MAAM,IAAI,GAAG,IAAI,GAAG,EAAU,CAAC;IAC/B,KAAK,MAAM,GAAG,IAAI,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,KAAK,CAAC,EAAE,CAAC;QACvC,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC;YAAE,SAAS;QAC7B,MAAM,EAAE,GAAG,OAAO,GAAG,CAAC,EAAE,KAAK,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC3D,IAAI,EAAE,CAAC,MAAM,KAAK,CAAC,IAAI,EAAE,CAAC,MAAM,GAAG,GAAG,IAAI,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAAE,SAAS;QACjE,MAAM,aAAa,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,CAAC,yBAAyB,CAAC;YAChE,CAAC,CAAC,GAAG,CAAC,yBAAyB,CAAC,GAAG,CAAC,CAAC,KAAK,EAAE,EAAE,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,eAAe,CAAC,CAAC,CAAC,KAAK,CAAC;YAC/F,CAAC,CAAC,EAAE,CAAC;QACP,MAAM,gBAAgB,GAAG,gBAAgB,CAAC,aAAa,CAAC,CAAC;QACzD,MAAM,iBAAiB,GAAG,eAAe,CAAC,GAAG,CAAC,sBAAsB,CAAC,CAAC;QACtE,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACb,MAAM,CAAC,IAAI,CAAC;YACV,EAAE;YACF,WAAW,EAAE,kBAAkB,CAAC,GAAG,CAAC,WAAW,EAAE,GAAG,CAAC,IAAI,EAAE;YAC3D,gBAAgB;YAChB,GAAG,CAAC,iBAAiB,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,aAAa,EAAE,iBAAiB,EAAE,CAAC;YAChF,GAAG,CAAC,OAAO,GAAG,CAAC,SAAS,KAAK,SAAS,CAAC,CAAC,CAAC,EAAE,SAAS,EAAE,GAAG,CAAC,SAAS,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SAC5E,CAAC,CAAC;IACL,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,qBAAqB,CACzC,SAAiB,EACjB,WAAyD,EACzD,WAAyB;IAEzB,MAAM,KAAK,GAAG,KAAK,CACjB,OAAO,EACP,CAAC,YAAY,EAAE,UAAU,EAAE,UAAU,EAAE,IAAI,EAAE,yBAAyB,CAAC,EACvE;QACE,KAAK,EAAE,CAAC,MAAM,EAAE,MAAM,EAAE,QAAQ,CAAC;QACjC,GAAG,EAAE,EAAE,GAAG,WAAW,EAAE;KACxB,CACF,CAAC;IACF,MAAM,KAAK,GAAG,eAAe,CAAC,EAAE,KAAK,EAAE,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC;IACvD,OAAO,MAAM,IAAI,OAAO,CAAsB,CAAC,aAAa,EAAE,YAAY,EAAE,EAAE;QAC5E,IAAI,OAAO,GAAG,KAAK,CAAC;QACpB,IAAI,KAAqB,CAAC;QAC1B,MAAM,gBAAgB,GAAG,GAAG,EAAE,CAAC,SAAS,CAAC;QACzC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC,OAAO,EAAE,gBAAgB,CAAC,CAAC;QAC1C,MAAM,iBAAiB,GAAG,KAAK,EAAE,YAAoB,EAAoB,EAAE;YACzE,IAAI,KAAK,CAAC,QAAQ,KAAK,IAAI,IAAI,KAAK,CAAC,UAAU,KAAK,IAAI;gBAAE,OAAO,IAAI,CAAC;YACtE,OAAO,MAAM,IAAI,OAAO,CAAU,CAAC,aAAa,EAAE,EAAE;gBAClD,MAAM,OAAO,GAAG,GAAG,EAAE;oBACnB,YAAY,CAAC,UAAU,CAAC,CAAC;oBACzB,aAAa,CAAC,IAAI,CAAC,CAAC;gBACtB,CAAC,CAAC;gBACF,MAAM,UAAU,GAAG,UAAU,CAAC,GAAG,EAAE;oBACjC,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;oBAC5B,aAAa,CAAC,KAAK,CAAC,CAAC;gBACvB,CAAC,EAAE,YAAY,CAAC,CAAC;gBACjB,UAAU,CAAC,KAAK,EAAE,EAAE,CAAC;gBACrB,KAAK,CAAC,IAAI,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;YAC/B,CAAC,CAAC,CAAC;QACL,CAAC,CAAC;QACF,MAAM,SAAS,GAAG,KAAK,IAAI,EAAE;YAC3B,KAAK,CAAC,kBAAkB,CAAC,MAAM,CAAC,CAAC;YACjC,KAAK,CAAC,KAAK,EAAE,CAAC;YACd,IAAI,CAAC;gBAAC,KAAK,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC,CAAC,iBAAiB,CAAC,CAAC;YACtD,IAAI,MAAM,iBAAiB,CAAC,EAAE,CAAC;gBAAE,OAAO;YACxC,IAAI,CAAC;gBAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC,CAAC,iBAAiB,CAAC,CAAC;YAC1D,IAAI,MAAM,iBAAiB,CAAC,GAAG,CAAC;gBAAE,OAAO;YACzC,IAAI,CAAC;gBAAC,KAAK,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC,CAAC,iBAAiB,CAAC,CAAC;YAC1D,IAAI,CAAC,MAAM,iBAAiB,CAAC,GAAG,CAAC,EAAE,CAAC;gBAClC,KAAK,CAAC,KAAK,CAAC,OAAO,EAAE,CAAC;gBACtB,KAAK,CAAC,MAAM,CAAC,OAAO,EAAE,CAAC;gBACvB,KAAK,CAAC,KAAK,EAAE,CAAC;YAChB,CAAC;QACH,CAAC,CAAC;QACF,IAAI,MAAkF,CAAC;QACvF,MAAM,YAAY,GAAG,GAAG,EAAE,CAAC,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,kCAAkC,CAAC,CAAC,CAAC;QACtF,MAAM,YAAY,GAAG,GAAG,EAAE,CAAC,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,kDAAkD,CAAC,CAAC,CAAC;QACtG,MAAM,OAAO,GAAG,GAAG,EAAE;YACnB,MAAM,KAAK,GAAG,IAAI,KAAK,CAAC,gDAAgD,CAAC,CAAC;YAC1E,KAAK,CAAC,IAAI,GAAG,YAAY,CAAC;YAC1B,KAAK,MAAM,CAAC,KAAK,CAAC,CAAC;QACrB,CAAC,CAAC;QACF,MAAM,GAAG,KAAK,EAAE,KAAwB,EAAE,SAA8B,EAAE,EAAE,EAAE;YAC5E,IAAI,OAAO;gBAAE,OAAO;YACpB,OAAO,GAAG,IAAI,CAAC;YACf,YAAY,CAAC,KAAK,CAAC,CAAC;YACpB,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;YACjC,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;YACjC,WAAW,EAAE,mBAAmB,CAAC,OAAO,EAAE,OAAO,CAAC,CAAC;YACnD,MAAM,SAAS,EAAE,CAAC;YAClB,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,OAAO,EAAE,gBAAgB,CAAC,CAAC;YAC3C,IAAI,KAAK,KAAK,SAAS;gBAAE,aAAa,CAAC,MAAM,CAAC,CAAC;;gBAC1C,YAAY,CAAC,KAAK,CAAC,CAAC;QAC3B,CAAC,CAAC;QACF,MAAM,YAAY,GAAG,CAAC,OAAgB,EAAE,EAAE;YACxC,IAAI,OAAO,IAAI,KAAK,CAAC,KAAK,CAAC,SAAS,IAAI,KAAK,CAAC,KAAK,CAAC,aAAa;gBAAE,OAAO;YAC1E,IAAI,CAAC;gBAAC,KAAK,CAAC,KAAK,CAAC,KAAK,CAAC,GAAG,IAAI,CAAC,SAAS,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC;YAAC,CAAC;YAAC,MAAM,CAAC,CAAC,0CAA0C,CAAC,CAAC;QACjH,CAAC,CAAC;QACF,KAAK,GAAG,UAAU,CAChB,GAAG,EAAE,CAAC,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,0CAA0C,CAAC,CAAC,EACxE,IAAI,CAAC,GAAG,CAAC,GAAG,EAAE,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,SAAS,CAAC,CAAC,CAC3C,CAAC;QACF,KAAK,CAAC,KAAK,EAAE,EAAE,CAAC;QAChB,KAAK,CAAC,IAAI,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;QAClC,KAAK,CAAC,IAAI,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;QAClC,WAAW,EAAE,gBAAgB,CAAC,OAAO,EAAE,OAAO,EAAE,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC;QAChE,KAAK,CAAC,EAAE,CAAC,MAAM,EAAE,CAAC,IAAI,EAAE,EAAE;YACxB,IAAI,OAAgB,CAAC;YACrB,IAAI,CAAC;gBACH,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;YAC7B,CAAC;YAAC,MAAM,CAAC;gBACP,OAAO;YACT,CAAC;YACD,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC;gBAAE,OAAO;YAC/B,IAAI,OAAO,OAAO,CAAC,MAAM,KAAK,QAAQ,IAAI,CAAC,OAAO,OAAO,CAAC,EAAE,KAAK,QAAQ,IAAI,OAAO,OAAO,CAAC,EAAE,KAAK,QAAQ,CAAC,EAAE,CAAC;gBAC7G,YAAY,CAAC;oBACX,EAAE,EAAE,OAAO,CAAC,EAAE;oBACd,KAAK,EAAE,EAAE,IAAI,EAAE,CAAC,KAAK,EAAE,OAAO,EAAE,2DAA2D,EAAE;iBAC9F,CAAC,CAAC;gBACH,OAAO;YACT,CAAC;YACD,IAAI,OAAO,CAAC,EAAE,KAAK,CAAC,IAAI,QAAQ,CAAC,OAAO,CAAC,MAAM,CAAC,EAAE,CAAC;gBACjD,YAAY,CAAC;oBACX,EAAE,EAAE,CAAC;oBACL,MAAM,EAAE,YAAY;oBACpB,MAAM,EAAE,EAAE,aAAa,EAAE,KAAK,EAAE,KAAK,EAAE,KAAK,EAAE;iBAC/C,CAAC,CAAC;gBACH,OAAO;YACT,CAAC;YACD,IAAI,OAAO,CAAC,EAAE,KAAK,CAAC;gBAAE,OAAO;YAC7B,IAAI,CAAC,QAAQ,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC;gBACrE,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,qDAAqD,CAAC,CAAC,CAAC;gBAC9E,OAAO;YACT,CAAC;YACD,KAAK,MAAM,CAAC,SAAS,EAAE,qBAAqB,CAAC,OAAO,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC;QACrE,CAAC,CAAC,CAAC;QACH,YAAY,CAAC;YACX,EAAE,EAAE,CAAC;YACL,MAAM,EAAE,YAAY;YACpB,MAAM,EAAE;gBACN,UAAU,EAAE,EAAE,IAAI,EAAE,YAAY,EAAE,KAAK,EAAE,YAAY,EAAE,OAAO,EAAE,GAAG,EAAE;gBACrE,YAAY,EAAE,EAAE,eAAe,EAAE,IAAI,EAAE;aACxC;SACF,CAAC,CAAC;QACH,IAAI,WAAW,EAAE,OAAO,KAAK,IAAI;YAAE,OAAO,EAAE,CAAC;IAC/C,CAAC,CAAC,CAAC;AACL,CAAC"}