@counterposition/skill-pi 0.80.3 → 0.84.2

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@counterposition/skill-pi",
-  "version": "0.80.3",
+  "version": "0.84.2",
   "description": "Pi coding agent reference skill — extensions, settings, providers, SDK/RPC, packages, and .pi/ config",
   "keywords": [
     "ai-skill",

package/skills/pi/SKILL.md

@@ -7,7 +7,7 @@ compatibility: "Pi-specific guidance for agents that support Agent Skills"
 
 # Pi Coding Agent
 
-Pi is a minimal terminal coding harness. Default tools: `read`, `write`, `edit`, `bash`, `grep`, `find`, `ls`. Modes: interactive (`pi`), print (`pi -p`), JSON (`--mode json`), RPC (`--mode rpc`), or embedded (`createAgentSession()`). Sub-agents, plan mode, permission flows, and MCP are intentionally left to extensions and Pi packages.
+Pi is a minimal terminal coding harness. Default tools: `read`, `write`, `edit`, `bash`, `grep`, `find`, `ls` (configurable via the `defaultTools` setting since Pi 0.84.2). Modes: interactive (`pi`), print (`pi -p`), JSON (`--mode json`), RPC (`--mode rpc`), or embedded (`createAgentSession()`). Sub-agents, plan mode, permission flows, and MCP are intentionally left to extensions and Pi packages. Skill synced with Pi 0.84.2.
 
 Pi's philosophy: **adapt Pi to your workflows, not the other way around**.
 
@@ -20,8 +20,10 @@ Core packages on npm (source: [github.com/earendil-works/pi](https://github.com/
 | `@earendil-works/pi-ai` | Unified LLM API across 20+ providers |
 | `@earendil-works/pi-agent-core` | Agent runtime with tool execution and state |
 | `@earendil-works/pi-tui` | Terminal UI components |
-| `@earendil-works/pi-coding-agent` | CLI, extensions, skills, sessions, settings |
+| `@earendil-works/pi-coding-agent` | CLI, extensions, skills, sessions, settings (also exports `./client` for remote sessions and `./rpc-entry`) |
 | `@earendil-works/pi-web-ui` | Web components for chat interfaces |
+| `@earendil-works/pi-client` / `@earendil-works/pi-protocol` | Experimental remote-session client and wire protocol (Pi 0.84.0) |
+| `@earendil-works/pi-telemetry` | Vendor-neutral telemetry contracts (Pi 0.84.0) |
 
 ## File System Layout
 
@@ -54,9 +56,9 @@ Core packages on npm (source: [github.com/earendil-works/pi](https://github.com/
 - **Settings** — Hierarchical JSON: project `.pi/settings.json` merges over global `~/.pi/agent/settings.json`. → `references/settings.md`
 - **Packages** — Bundles of extensions/skills/prompts/themes via npm, git, or local paths. Installed with `pi install`. → `references/packages.md`
 - **Project trust** — Pi 0.79+ asks before loading project-local settings, resources, and packages; decisions persist in `~/.pi/agent/trust.json`. `--approve`/`--no-approve` override per run; `defaultProjectTrust` sets the non-interactive fallback. → `references/settings.md`
-- **Context files & prompt templates** — Pi loads `AGENTS.md` / `CLAUDE.md` from the agent dir and from `cwd` up through ancestors. `.pi/SYSTEM.md` replaces the system prompt; `APPEND_SYSTEM.md` appends. Prompt templates in `prompts/` become slash commands. → `references/settings.md`
+- **Context files & prompt templates** — Pi loads `AGENTS.md` / `CLAUDE.md` from the agent dir and from `cwd` up through ancestors. Per-directory `AGENTS.override.md` (Pi 0.84.0) replaces same-directory context files while others layer normally. `.pi/SYSTEM.md` replaces the system prompt; `APPEND_SYSTEM.md` appends. Prompt templates in `prompts/` become slash commands. → `references/settings.md`
 - **SDK** — Programmatic embedding via `createAgentSession()`; `createAgentSessionRuntime()` for session replacement. → `references/sdk.md`
-- **Custom providers & models** — `models.json` or extension `pi.registerProvider()` for any OpenAI-/Anthropic-/Google-compatible or custom LLM endpoint. → `references/providers.md`
+- **Custom providers & models** — `models.json` or extension `pi.registerProvider()` (config form or complete pi-ai providers) for any OpenAI-/Anthropic-/Google-compatible or custom LLM endpoint. → `references/providers.md`
 - **Sessions & compaction** — Append-only JSONL with a tree structure; branch with `/tree`, `/fork`, `/clone`, compact with `/compact`. Auto-compaction triggers when `contextTokens > contextWindow - reserveTokens`. Extensions intercept via `session_before_compact`. → `references/extensions.md` and `references/sdk.md`
 
 ## How to Use This Skill

package/skills/pi/references/extensions.md

@@ -23,7 +23,7 @@ import { StringEnum } from "@earendil-works/pi-ai";
 import { Text } from "@earendil-works/pi-tui";
 ```
 
-- Use `typebox` (1.x) for schemas.
+- Use `typebox` (1.x) for schemas. Pi 0.83.0 bundles TypeBox 1.3.7, which removed deprecated APIs (`Type.Base`, `Type.Awaited`, `Type.Promise`, `Type.AsyncIterator`, `Type.Iterator`, `Type.Options`, `Value.Mutate`) — migrate extension sources off those before typechecking against current Pi.
 - Use `StringEnum` from `@earendil-works/pi-ai` for Google-compatible string enums.
 - Pi 0.80.0 moved pi-ai's old global API (`getModel`, `getModels`, `stream`, `complete`, `registerApiProvider`, ...) off the pi-ai root entrypoint to `@earendil-works/pi-ai/compat`. Extensions keep working unchanged at runtime (the extension loader aliases the root to the compat superset), but extension sources that typecheck against pi-ai's published types must import those APIs from `@earendil-works/pi-ai/compat`. `StringEnum` and the type surface remain on the root. The compat entrypoint and loader alias will be removed in a future release.
 
@@ -90,13 +90,15 @@ Useful agent events:
 - `input` — carries `event.streamingBehavior` (`"steer" | "followUp" | undefined`) so handlers can distinguish idle prompts, mid-stream steers, and queued follow-ups; return `{ action: "transform" | "handled" | "continue" }`
 - `before_agent_start` — receives `event.systemPromptOptions` (a `BuildSystemPromptOptions`) so handlers can inspect the structured inputs feeding the system prompt
 - `agent_start`
-- `agent_end`
+- `agent_end` — one low-level run; Pi may still auto-retry, auto-compact and retry, or continue with queued follow-ups
+- `agent_settled` (Pi 0.80.4) — fired when no retry/compaction/follow-up remains; use for status integrations that need to know Pi will not continue automatically (`ctx.isIdle()` is true here unless another extension started a new run)
 - `turn_start`
 - `turn_end`
 - `message_start`
 - `message_update`
 - `message_end` — return a replacement message to override usage/cost or rewrite the finalized assistant message
 - `context`
+- `before_provider_headers` (Pi 0.80.4) — mutate `event.headers` in place after outgoing HTTP headers are assembled: set a key to a string to add/override, `null` to delete. Runs once per provider request; retries reuse the same headers
 - `before_provider_request`
 - `after_provider_response` — inspect the provider HTTP status and headers before stream consumption
 - `model_select`
@@ -110,7 +112,7 @@ Useful tool events:
 - `tool_result`
 - `tool_execution_end`
 
-Tool results may include `terminate: true` to end the current tool batch without an automatic follow-up LLM turn — useful for tools that produce a structured final answer (see `examples/extensions/structured-output.ts` in the Pi repo).
+Tool results may include `terminate: true` to end the current tool batch without an automatic follow-up LLM turn — useful for tools that produce a structured final answer (see `examples/extensions/structured-output.ts` in the Pi repo). Since Pi 0.84.1, blocked `tool_call` results may also return `terminate: true`; the agent stops early only when every finalized result in the batch is terminating. Tools that make nested LLM calls can return their combined `Usage` as `usage` on the tool result — Pi persists it and includes it in footer, `/session`, and RPC session totals (Pi 0.81.0), and `tool_result` handlers may inspect or replace that value.
 
 `session_shutdown` events carry `event.reason` (`"quit" | "reload" | "new" | "resume" | "fork"`) and, where applicable, `event.targetSessionFile` so cleanup logic can distinguish teardown paths.
 
@@ -121,8 +123,8 @@ Tool results may include `terminate: true` to end the current tool batch without
 - `ctx.ui` for interactive UI hooks
 - `ctx.mode` — `"tui" | "rpc" | "json" | "print"`; gate terminal-only features on `ctx.mode === "tui"`
 - `ctx.cwd`
-- `ctx.sessionManager` (read-only)
-- `ctx.modelRegistry` / `ctx.model`
+- `ctx.sessionManager` (read-only; `buildContextEntries()` returns active-branch entries with compaction applied, Pi 0.80.4)
+- `ctx.modelRegistry` / `ctx.model` / `ctx.thinkingLevel` / `ctx.scopedModels` (Pi 0.83.0) — `ctx.modelRegistry` is the synchronous extension-facing facade; its `refresh()` became `Promise<void>` in Pi 0.80.8 (await it before synchronous registry reads). Since Pi 0.84.x it also exposes `getProvider(id)` (effective pi-ai provider) and `getProviderAuth(id)` (API key, headers, base URL, provider-scoped env — no loaded model required). `ctx.scopedModels` is the read-only session model scope resolved from `--models`/`enabledModels` (`{ model, thinkingLevel? }[]`, empty when unscoped); prefer it over enumerating `getAvailable()` for model pickers
 - `ctx.hasUI` — `true` in TUI and RPC modes
 - `ctx.signal` — the active agent abort signal (or `undefined` when idle); pass it to `fetch`/model calls for abort-aware nested work
 - `ctx.isIdle()` / `ctx.hasPendingMessages()`
@@ -171,6 +173,42 @@ Important rules:
 4. Use the file-mutation queue for write/edit style tools.
 5. Put reconstructable state in `details`; it persists in session history.
 
+## Dynamic Tool Loading
+
+Pi 0.80.7 lets an extension register many tools while keeping only a small initial set active, then add more during execution — cache-friendly on models with native deferred loading. Lifecycle:
+
+1. Register every tool with `pi.registerTool()` (all appear in `pi.getAllTools()`).
+2. Keep a loader tool (e.g. `search_tools`) active; leave searchable tools inactive — e.g. on `session_start`, `pi.setActiveTools()` to the filtered set.
+3. During loader execution, call `pi.setActiveTools([...pi.getActiveTools(), ...matches])`. The change must be purely additive; unknown names are ignored.
+4. Pi records the added tools on that tool result and exposes their definitions before the next model response.
+
+Native deferred loading preserves the cached prompt prefix on Anthropic Sonnet/Opus/Fable ≥ 4.5 (not Haiku) and OpenAI `gpt-5.4`+; Kimi K3 works via `compat.deferredToolsMode: "kimi"` (Pi 0.80.9). For verified custom models/proxies, enable `compat.supportsToolReferences: true` (`anthropic-messages`) or `compat.supportsToolSearch: true` (`openai-responses`/`openai-codex-responses`). All other models fall back to sending the full active tool list on the next request — activation still works, but may invalidate the provider's cached prefix. Non-additive changes (removals/replacements) always use the fallback.
+
+Cache tips: keep the loader active for the whole session; add rather than replace. Activating a tool that has `promptSnippet`/`promptGuidelines` rebuilds the system prompt and can invalidate the prefix even with native support — lazily loaded tools should rely on their `description` alone.
+
+## Constrained Tool Sampling
+
+`ToolDefinition.constrainedSampling` (Pi 0.82.0) lets a tool prefer or require provider-side output constraints:
+
+- `{ type: "json_schema", strict: "prefer" | "require" }` — strict JSON-schema constrained decoding of the tool's own TypeBox parameters
+- `{ type: "grammar", variants: { openai_lark?: string, openai_regex?: string } }` — OpenAI Lark/regex grammar encodings
+- Set `false` to explicitly disable for that tool
+
+Model capability metadata gates what is actually sent: `compat.supportsStrictTools` / Anthropic built-ins enable strict JSON-schema tools, and `compat.supportsOpenAIGrammarTools` marks endpoints that accept Lark/regex grammar tools (enabled in generated metadata for GPT-5+ across OpenAI, Codex, Azure, GitHub Copilot, opencode, Cloudflare AI Gateway). Unsupported combinations fall back to normal function tools rather than failing. Independently, Pi 0.84.2 ships experimental strict JSON-schema sampling for its default `read`/`bash`/`edit`/`write` tools under `PI_EXPERIMENTAL=1`.
+
+## Display Transformers
+
+`pi.registerMarkdownTransformer(transformer)` (Pi 0.84.0) chains display-only Markdown transforms over user text, assistant text, and thinking blocks:
+
+```typescript
+pi.registerMarkdownTransformer((markdown, { messageType, isStreaming }) => {
+  if (isStreaming || messageType === "assistant-thinking") return markdown;
+  return markdown.replaceAll("-->", "→");
+});
+```
+
+Context carries `messageType` (`"user" | "assistant" | "assistant-thinking"`), `isStreaming`, and `availableWidth` (terminal columns). The hook runs for new messages, restored sessions, and terminal resizes, so keep it synchronous and cheap; a throwing transformer leaves prior output intact and Pi continues the chain. The session and model context are never modified — this is render-only.
+
 ## UI Methods
 
 Check `ctx.hasUI` first. Non-interactive modes may not support UI.
@@ -240,6 +278,15 @@ Store extension state with custom session entries:
 pi.appendEntry("my-extension-state", { key: "value" });
 ```
 
+Custom entries never participate in LLM context. Since Pi 0.80.4 they can also render in the interactive transcript via `pi.registerEntryRenderer(customType, renderer)` — the TUI-only counterpart to `pi.registerMessageRenderer()` (whose custom messages DO enter LLM context via `pi.sendMessage()`):
+
+```typescript
+pi.registerEntryRenderer("status-card", (entry, { expanded }, theme) => {
+  return new Text(theme.fg("accent", JSON.stringify(entry.data)));
+});
+pi.appendEntry("status-card", { title: "Indexed files", count: 17 });
+```
+
 Use lifecycle hooks to restore it:
 
 ```typescript
@@ -280,12 +327,18 @@ pi.registerProvider("anthropic", { baseUrl: "https://proxy.example.com" });
 pi.unregisterProvider("my-provider");
 ```
 
+Dynamic providers can implement `refreshModels(context)` (Pi 0.80.8) for model discovery: Pi calls it during catalog refresh (`/model`, `pi update --models`) and publishes the returned list; its models replace extension-provided `models`. Since Pi 0.84.0 the context is read-snapshot + generation-checked publication: read `context.stored` (the persisted provider snapshot) instead of `context.store`, and persist through `context.publish({ update?, persist? })` — `persist` omitted leaves storage unchanged, a `ModelsStoreEntry` writes it, `persist: null` deletes it. Config-form callbacks that only return models remain unchanged.
+
+Since Pi 0.81.0, extensions can also register a complete pi-ai `Provider` via `pi.registerProvider(createProvider({...}))` — native auth (`login`/`resolve`), `getModels`, `refreshModels`, `filterModels`, and custom streaming. The registered provider becomes the composition base; `models.json` overrides still apply above it. Prefer this when you need real authentication or stream behavior rather than just a model list. The legacy config form (`name`, `baseUrl`, `apiKey`, `models`, `oauth`, `streamSimple`) remains supported. OAuth `refreshToken(credentials, signal)` callbacks must accept and honor the abort signal since Pi 0.84.0.
+
 If you need auth for a specific model request, use:
 
 ```typescript
 const auth = await ctx.modelRegistry.getApiKeyAndHeaders(model);
 ```
 
+(Still supported for extensions in Pi 0.80.8+; SDK code should use `ModelRuntime.getAuth()` — see `references/sdk.md`. `ProviderHeaders` values are `string | null` since Pi 0.84.0, where `null` is a delete marker: handle `null` when inspecting, pass through unchanged when forwarding to pi-ai streams.)
+
 Pass `shouldStopAfterTurn` via the SDK to exit the agent loop gracefully after a completed turn. See `references/sdk.md` and `references/providers.md` for the full provider/model schema.
 
 ## Handy Patterns

package/skills/pi/references/packages.md

@@ -19,9 +19,11 @@ pi list
 pi update                 # update pi only (changed in Pi 0.79.7)
 pi update --all           # update pi + packages, reconcile pinned git refs
 pi update --extensions    # update packages only
+pi update --models        # refresh model catalogs only (Pi 0.80.8)
 pi update --self          # update pi only
 pi update <source>        # update one package
-pi config
+pi config                 # starts in global settings; Tab switches scope
+pi config -l              # start in project overrides (.pi/settings.json)
 ```
 
 By default, `install` and `remove` write to user settings (`~/.pi/agent/settings.json`). Use `-l` to write to project settings (`.pi/settings.json`) instead. `pi update` installs the exact version returned by the update check.
@@ -99,6 +101,7 @@ If there is no `pi` manifest, Pi auto-discovers:
   - `@earendil-works/pi-coding-agent`
   - `@earendil-works/pi-tui`
   - `typebox`
+- Pin dev dependencies to a current Pi line (e.g. `^0.84.2`) so typecheck/tests validate against what users run; keep peers `"*"` so installs stay compatible across Pi versions
 - Other Pi packages must be bundled explicitly
 - Git package installs run `npm install --omit=dev` — runtime needs go in `dependencies`, not `devDependencies`
 
@@ -129,6 +132,6 @@ Rules:
 
 ## Scope and Deduplication
 
-- Project settings override global settings for the same package identity
+- Project settings override global settings for the same package identity — unless the project entry has `autoload: false`, in which case it is applied as a delta over the global entry (Pi 0.80.4)
 - Identity is package name for npm, repo URL for git, and resolved absolute path for local sources
-- `pi config` can enable or disable package resources after installation
+- `pi config` can enable or disable package resources after installation; since Pi 0.80.4 it manages global vs project-local scopes (Tab to switch, `pi config -l` to start in project mode with inherited global resources dimmed)

package/skills/pi/references/providers.md

@@ -9,8 +9,13 @@ Use `/login` in interactive mode, then select a provider. The `/login` selector
 - Anthropic Claude Pro / Max — third-party usage draws from extra usage and is billed per token (suppress the warning via `warnings.anthropicExtraUsage`)
 - OpenAI ChatGPT Plus / Pro (Codex) — `/login` defaults to browser auth but can use a device-code flow for headless environments
 - GitHub Copilot
+- OpenRouter (Pi 0.82.0) — `/login openrouter` runs a PKCE flow that mints a user-controlled API key billed from OpenRouter credits; on headless/SSH hosts paste the redirect URL or authorization code into the prompt (Pi 0.83.0). `OPENROUTER_API_KEY` still works via **Use an API key**
+- xAI (Grok/X subscription, Pi 0.80.8) — `/login xai` then **Use a subscription** (device-code OAuth); `XAI_API_KEY` remains available via **Use an API key**
+- Radius (Pi 0.80.8) — a dynamic `pi-messages` gateway; `/login radius` stores OAuth tokens, and custom Radius gateways can be declared in `models.json` with `"oauth": "radius"` plus a gateway `baseUrl`
 
-Use `/logout` to clear stored OAuth credentials. Pi 0.71.0 removed built-in Google Gemini CLI and Google Antigravity providers.
+`/login <provider>` with autocomplete works since Pi 0.80.4, and login methods are provider-owned since 0.80.8 (registered pi-ai providers expose their own auth options and status in `/login`). Use `/logout` to clear stored OAuth credentials. Pi 0.71.0 removed built-in Google Gemini CLI and Google Antigravity providers. `pi auth check [provider|model]` (Pi 0.84.1) preflights credential readiness; `pi auth print-api-key` / `pi auth print-bearer-token` (Pi 0.83.0) export resolved credentials to external clients, refreshing OAuth as needed.
+
+Since Pi 0.80.8, built-in catalogs are complemented by dynamic ones: `/model` refreshes configured providers in the background, `pi update --models` forces an immediate refresh, and refreshed catalogs are cached in `~/.pi/agent/models-store.json` for offline use. Refreshes revalidate with `If-None-Match` since Pi 0.82.1 so unchanged catalogs answer `304`.
 
 ## API Key Providers
 
@@ -35,12 +40,17 @@ Use `/logout` to clear stored OAuth credentials. Pi 0.71.0 removed built-in Goog
 | ZAI Coding Plan (China) | `ZAI_CODING_CN_API_KEY` | `zai-coding-cn` |
 | OpenCode Zen | `OPENCODE_API_KEY` | `opencode` |
 | OpenCode Go | `OPENCODE_API_KEY` | `opencode-go` |
+| Amazon Bedrock | `AWS_BEARER_TOKEN_BEDROCK` | `amazon-bedrock` |
+| Radius | `RADIUS_API_KEY` | `radius` |
 | Hugging Face | `HF_TOKEN` | `huggingface` |
 | Fireworks | `FIREWORKS_API_KEY` | `fireworks` |
 | Together AI | `TOGETHER_API_KEY` | `together` |
+| Baseten | `BASETEN_API_KEY` | `baseten` |
 | Kimi For Coding | `KIMI_API_KEY` | `kimi-coding` |
 | MiniMax | `MINIMAX_API_KEY` | `minimax` |
 | MiniMax (China) | `MINIMAX_CN_API_KEY` | `minimax-cn` |
+| Qwen Token Plan / Individual | `QWEN_TOKEN_PLAN_API_KEY` | `qwen-token-plan` / `qwen-token-plan-individual` (Pi 0.81.0 / 0.84.1) |
+| Qwen Token Plan (China) | `QWEN_TOKEN_PLAN_CN_API_KEY` | `qwen-token-plan-cn` (Pi 0.81.0) |
 | Xiaomi MiMo | `XIAOMI_API_KEY` | `xiaomi` |
 | Xiaomi MiMo Token Plan (CN/AMS/SGP) | `XIAOMI_TOKEN_PLAN_{CN,AMS,SGP}_API_KEY` | `xiaomi-token-plan-{cn,ams,sgp}` |
 
@@ -99,7 +109,7 @@ export AZURE_OPENAI_DEPLOYMENT_NAME_MAP=gpt-4o=my-gpt4o
 
 ### Amazon Bedrock
 
-Pi supports standard AWS auth flows:
+`/login amazon-bedrock` stores a Bedrock API key (Pi 0.80.7). Ambient AWS credential flows also work — these keep using SigV4 authentication:
 
 - `AWS_PROFILE`
 - `AWS_ACCESS_KEY_ID` / `AWS_SECRET_ACCESS_KEY`
@@ -182,15 +192,24 @@ Common `api` values:
 
 ### Model & Compatibility Knobs
 
-- `thinkingLevelMap` (Pi 0.72) replaces `compat.reasoningEffortMap`. Map pi levels (`off`/`minimal`/`low`/`medium`/`high`/`xhigh`) to provider values; use `null` to hide a level.
+- `samplingParams` (Pi 0.84.0) is a free-form object merged verbatim into every request body after pi's own fields (its keys win) — for OpenAI-compatible APIs only (`openai-completions`, `openai-responses`, `azure-openai-responses`). Use it for server-specific knobs like llama.cpp `min_p` or vLLM `top_k`; in `modelOverrides` it merges per key.
+- `thinkingLevelMap` (Pi 0.72) replaces `compat.reasoningEffortMap`. Map pi levels (`off`/`minimal`/`low`/`medium`/`high`/`xhigh`/`max`) to provider values; use `null` to hide a level. Maps may contain holes (Pi 0.80.6) — e.g. expose `high` and `max` without `xhigh`. When a key is omitted, standard levels through `high` use the provider default mapping, but the extended `xhigh`/`max` levels are unsupported.
+- `cost` supports request-wide input pricing tiers (Pi 0.80.6): a `tiers` array where each tier supplies a complete alternate rate set and applies to the whole request when total input usage (`input + cacheRead + cacheWrite`) exceeds `inputTokensAbove`; the highest matching threshold wins. Also usable in `modelOverrides` and extension-registered providers.
+- `modelOverrides` applies to built-in and (since Pi 0.80.4) extension-registered provider models; per-model fields: `name`, `reasoning`, `thinkingLevelMap`, `input`, `cost` (partial), `contextWindow`, `maxTokens`, `samplingParams`, `headers`, `compat`.
 - `openRouterRouting` is forwarded as-is in the OpenRouter `provider` field (fallbacks, ZDR, ignore lists, throughput/latency).
-- `compat.thinkingFormat` supports OpenAI-compatible reasoning variants: `openrouter` sends `reasoning: { effort }`, `together` sends `reasoning: { enabled }` plus `reasoning_effort` when supported, `qwen-chat-template` targets local Qwen-compatible servers that read `chat_template_kwargs.enable_thinking`, and `chat-template` (Pi 0.79.9) sends configurable `chat_template_kwargs` via `compat.chatTemplateKwargs` — e.g. `{ "thinking": { "$var": "thinking.enabled" } }` for DeepSeek models behind vLLM/Hugging Face chat templates (`"$var"` accepts `"thinking.enabled"` or `"thinking.effort"`).
-- Advanced `compat` flags exist for proxy quirks (`cacheControlFormat`, `supportsReasoningEffort`, `supportsLongCacheRetention`, `supportsEagerToolInputStreaming`, `sendSessionIdHeader`, `sendSessionAffinityHeaders`). See [Pi `docs/models.md`](https://github.com/earendil-works/pi/blob/main/packages/coding-agent/docs/models.md) when a proxy rejects pi's defaults.
+- `compat.thinkingFormat` supports OpenAI-compatible reasoning variants: `openrouter` sends `reasoning: { effort }`, `together` sends `reasoning: { enabled }` plus `reasoning_effort` when supported, `qwen-chat-template` targets local Qwen-compatible servers that read `chat_template_kwargs.enable_thinking`, `chat-template` (Pi 0.79.9) sends configurable `chat_template_kwargs` via `compat.chatTemplateKwargs` — e.g. `{ "thinking": { "$var": "thinking.enabled" } }` for DeepSeek models behind vLLM/Hugging Face chat templates — and `baseten` (Pi 0.84.0) sends `chat_template_args` via `compat.chatTemplateArgs`.
+- **Breaking (Pi 0.80.7):** `compat.sendSessionIdHeader` was removed. Session affinity is now controlled by `compat.sessionAffinityFormat` (`"openai"` sends `session_id`/`x-client-request-id`, `"openai-nosession"` omits the underscore-containing `session_id` header, `"openrouter"` sends `x-session-id`; default auto-detected). Replace `sendSessionIdHeader: false` with `sessionAffinityFormat: "openai-nosession"`. `sendSessionAffinityHeaders` still gates the behavior for `openai-completions`.
+- `compat.deferredToolsMode: "kimi"` (Pi 0.80.9) enables Kimi's deferred tool serialization; `compat.supportsToolReferences` / `compat.supportsToolSearch` enable native dynamic tool loading on verified custom endpoints, and `compat.supportsStrictTools` / `compat.supportsOpenAIGrammarTools` / per-model `constrainedSampling` gate constrained sampling (see `references/extensions.md`). `compat.supportsFinishReason: false` (Pi 0.84.0) tells pi to infer stop/toolUse for OpenAI-compatible streams that omit `finish_reason`.
+- Advanced `compat` flags exist for proxy quirks (`cacheControlFormat` — now also covering tool-result text content, `supportsReasoningEffort`, `supportsLongCacheRetention`, `supportsEagerToolInputStreaming`, `supportsStrictMode`). See [Pi `docs/models.md`](https://github.com/earendil-works/pi/blob/main/packages/coding-agent/docs/models.md) when a proxy rejects pi's defaults.
 
 ### Context Overflow Recovery
 
 Pi can auto-compact and retry when a provider fails with a recognized context-window overflow. For custom providers whose overflow text Pi does not recognize, use a provider-scoped `message_end` extension handler to rewrite the finalized assistant `errorMessage` so it starts with `context_length_exceeded`.
 
+### llama.cpp
+
+Pi supports the llama.cpp router server (Pi 0.81.0): configure with `/login llama.cpp`, search/download Hugging Face models and load/unload with live progress via `/llama`, then select loaded models in `/model`. The model catalog persists across restarts (fixed in Pi 0.82.1). See [Pi `docs/llama-cpp.md`](https://github.com/earendil-works/pi/blob/main/packages/coding-agent/docs/llama-cpp.md).
+
 ## Custom Providers via Extensions
 
 Use extensions when you need custom streaming, custom headers, OAuth/device-flow logic, or want to override an existing provider's `baseUrl`/`headers`:
@@ -219,7 +238,7 @@ pi.registerProvider("anthropic", { baseUrl: "https://proxy.example.com" });
 
 ## SDK / Extension Auth Lookup
 
-If extension or SDK code needs auth for a specific model request, use `getApiKeyAndHeaders(model)` rather than the removed `getApiKey(model)`:
+If extension code needs auth for a specific model request, use `getApiKeyAndHeaders(model)` rather than the removed `getApiKey(model)`:
 
 ```typescript
 const auth = await ctx.modelRegistry.getApiKeyAndHeaders(model);
@@ -228,4 +247,4 @@ if (!auth.ok) throw new Error(auth.error);
 const { apiKey, headers } = auth;
 ```
 
-This matters for providers whose headers or auth values resolve dynamically on every request.
+This matters for providers whose headers or auth values resolve dynamically on every request. In SDK code, use `ModelRuntime.getAuth(providerOrModel)` instead (Pi 0.80.8) — passing a model also resolves built-in, `models.json`, and extension model headers; see `references/sdk.md`.

package/skills/pi/references/sdk.md

@@ -12,19 +12,16 @@ npm install @earendil-works/pi-coding-agent
 
 ```typescript
 import {
-  AuthStorage,
   createAgentSession,
-  ModelRegistry,
+  ModelRuntime,
   SessionManager,
 } from "@earendil-works/pi-coding-agent";
 
-const authStorage = AuthStorage.create();
-const modelRegistry = ModelRegistry.create(authStorage);
+const modelRuntime = await ModelRuntime.create();
 
 const { session } = await createAgentSession({
   sessionManager: SessionManager.inMemory(),
-  authStorage,
-  modelRegistry,
+  modelRuntime,
 });
 
 session.subscribe((event) => {
@@ -38,17 +35,16 @@ await session.prompt("What files are in the current directory?");
 
 Run with `npx tsx my-script.ts`.
 
-**Important:** `ModelRegistry` no longer has a public constructor. Use `ModelRegistry.create(authStorage, modelsJsonPath?)` for file-backed registries or `ModelRegistry.inMemory(authStorage)` for built-in models only.
+**Breaking (Pi 0.80.8):** `ModelRuntime` replaced the old `AuthStorage`/`ModelRegistry` pair as the SDK model/auth facade. `CreateAgentSessionOptions.authStorage` and `modelRegistry` are gone — pass the async `modelRuntime` instead. `AuthStorage` is no longer exported; use `ModelRuntime` (or a custom pi-ai `CredentialStore`), or `readStoredCredential()` for one-off reads of `auth.json`. `ModelRegistry` still exists only as the synchronous extension-facing compatibility facade, and its `refresh()` is now `Promise<void>`.
 
 ## Key Imports
 
 ```typescript
 import {
-  AuthStorage,
   createAgentSession,
   DefaultResourceLoader,
   defineTool,
-  ModelRegistry,
+  ModelRuntime,
   SessionManager,
 } from "@earendil-works/pi-coding-agent";
 import { getBuiltinModel } from "@earendil-works/pi-ai/providers/all";
@@ -60,16 +56,14 @@ import { getBuiltinModel } from "@earendil-works/pi-ai/providers/all";
 
 ```typescript
 const cwd = "/path/to/project";
-const authStorage = AuthStorage.create();
-const modelRegistry = ModelRegistry.create(authStorage);
+const modelRuntime = await ModelRuntime.create();
 
 const { session } = await createAgentSession({
   cwd,
   agentDir: "~/.pi/agent",
-  authStorage,
-  modelRegistry,
+  modelRuntime,
   model: getBuiltinModel("anthropic", "claude-opus-4-8"),
-  thinkingLevel: "medium",
+  thinkingLevel: "medium",  // off | minimal | low | medium | high | xhigh | max
   scopedModels: [
     { model: getBuiltinModel("anthropic", "claude-opus-4-8"), thinkingLevel: "high" },
   ],
@@ -90,6 +84,52 @@ Notes:
 - `DefaultResourceLoader` loads extensions, skills, prompt templates, themes, and context files. Replace it to drive resource discovery from custom sources (and it must implement `loadProjectContextFiles()` if you want `AGENTS.md`/`CLAUDE.md` discovery; that helper is also exported standalone).
 - Pass `shouldStopAfterTurn(state) => boolean` (Pi 0.72.0) to exit the agent loop gracefully after a completed turn.
 
+## Models & Auth (`ModelRuntime`)
+
+`ModelRuntime` (Pi 0.80.8) implements the pi-ai `Models` interface and owns credential storage. Auth resolution priority: runtime overrides (`setRuntimeApiKey`, not persisted) → stored credentials in `auth.json` → environment variables → fallback resolver for `models.json` keys.
+
+```typescript
+import { InMemoryCredentialStore } from "@earendil-works/pi-ai";
+import { ModelRuntime } from "@earendil-works/pi-coding-agent";
+
+// Default: ~/.pi/agent/auth.json and ~/.pi/agent/models.json
+const modelRuntime = await ModelRuntime.create();
+
+// Model lookup (use pi-ai Models methods directly; the old
+// getAll()/find()/getSnapshot() projections were removed)
+const model = modelRuntime.getModel("my-provider", "my-model");
+const available = await modelRuntime.getAvailable();  // only models with valid auth
+
+// Provider-owned auth methods and current status
+for (const provider of modelRuntime.getProviders()) {
+  const status = await modelRuntime.checkAuth(provider.id);
+}
+
+modelRuntime.setRuntimeApiKey("anthropic", "sk-my-temp-key");
+
+// Custom locations, or inject any pi-ai CredentialStore
+const custom = await ModelRuntime.create({ authPath: "/my/auth.json", modelsPath: "/my/models.json" });
+const inMemory = await ModelRuntime.create({ credentials: new InMemoryCredentialStore() });
+```
+
+`ModelRuntime.getAuth(providerOrModel)` assembles final request auth (replacing `ModelRegistry.getApiKeyAndHeaders()` on the SDK side); passing a model also resolves built-in, `models.json`, and extension model headers. Dynamic provider catalogs refresh via async `ModelRuntime.refresh()` and are cached in `~/.pi/agent/models-store.json`.
+
+Pi 0.84.x made catalog refresh cancellation-aware: `refresh(options?)` accepts `{ providers?, allowNetwork?, force?, signal? }` and returns a `ModelsRefreshResult` (`{ errors, aborted }`) instead of discarding provider errors; `create({ signal })`, `login()`, `logout()`, and the now-async `setRuntimeApiKey()/removeRuntimeApiKey()` accept optional abort signals too (unbounded when omitted — applications own deadline policy). Credential mutations resolve once local state is consistent, without waiting for remote freshness; if that synchronization fails they reject with the exported `CredentialSynchronizationError` (inspect `providerId`, `operation`, `credential`, `cause` rather than retrying blindly). A failed refresh never undoes a committed credential change, and each `refresh()` starts a fresh generation so stale refreshes cannot publish afterward.
+
+To match CLI model parsing, use the exported resolver helpers (Pi 0.80.4):
+
+```typescript
+import { resolveCliModel, resolveModelScopeWithDiagnostics } from "@earendil-works/pi-coding-agent";
+
+const cliModel = resolveCliModel({ cliModel: "anthropic/claude-opus-4-8:high", modelRuntime });
+const { scopedModels, diagnostics } = await resolveModelScopeWithDiagnostics(
+  ["anthropic/*:high", "gpt-5"],
+  modelRuntime,
+);
+```
+
+`resolveCliModel()` resolves against all registered models (so `--api-key`-style first-time setup works before stored auth exists); `resolveModelScopeWithDiagnostics()` matches `--models`/`enabledModels` semantics and returns warnings instead of printing.
+
 ## Prompting & Queueing
 
 ```typescript
@@ -193,7 +233,8 @@ session.subscribe((event) => {
     case "tool_execution_update":
     case "tool_execution_end":
     case "agent_start":
-    case "agent_end":
+    case "agent_end":      // one low-level run; may be followed by retry/compaction/queued follow-ups
+    case "agent_settled":  // fully settled — no automatic continuation left (Pi 0.80.4)
     case "turn_start":
     case "turn_end":
     case "queue_update":
@@ -201,18 +242,30 @@ session.subscribe((event) => {
     case "compaction_end":
     case "auto_retry_start":
     case "auto_retry_end":
+    case "summarization_retry_scheduled":   // Pi 0.81.1 — compaction/branch-summary retries
+    case "summarization_retry_attempt_start":
+    case "summarization_retry_finished":
       break;
   }
 });
 ```
 
-`compaction_end` results and RPC `compact` responses include estimated post-compaction token counts (Pi 0.79.8) so clients can show the approximate context reduction.
+`compaction_end` results and RPC `compact` responses include estimated post-compaction token counts (Pi 0.79.8) so clients can show the approximate context reduction. Since Pi 0.81.0, summaries persist their LLM `usage`, and session token/cost totals (footer, `/session`, RPC) include assistant messages, tool-reported usage, and summary generation.
+
+**Streaming deltas (Pi 0.84.x, JSON/RPC wire only):** `message_update` records are delta-only — the cumulative `message` field and `assistantMessageEvent.partial` were removed to keep stream size linear. A top-level `usage` carries latest cumulative provider usage; assemble partial text from `contentIndex` + `delta` between `message_start` and `message_end`, which remains authoritative. In-process SDK subscribers still receive full events.
+
+## Remote Sessions (experimental)
+
+Pi 0.84.0 added transport-neutral remote-session client APIs: `@earendil-works/pi-client` (`PiClient`) with a CBOR protocol over Unix sockets, `@earendil-works/pi-protocol` for the wire contract, and `@earendil-works/pi-coding-agent/client` exporting a `RemoteSession` controller with transcript reducers. Sessions list durable `SessionMetadata`; runtime phase/model/lock state lives on acquired `SessionSnapshot` values. Expect these APIs to evolve while experimental.
 
 ## RPC Mode
 
 `pi --mode rpc` speaks newline-delimited JSON over stdio. Additions since Pi 0.79:
 
-- `get_entries` / `get_tree` (Pi 0.80.3) read session entries and tree snapshots over RPC.
+- `agent_settled` event (Pi 0.80.4) fires when a run is fully settled — no automatic retry, compaction retry, or queued continuation remains; `agent_end` now carries `willRetry`. `set_thinking_level` accepts `"max"` where the model supports it.
+- `get_entries` / `get_tree` (Pi 0.80.3) read session entries and tree snapshots over RPC; `get_available_thinking_levels` (Pi 0.81.0) lists the current model's supported thinking levels.
+- Direct RPC `bash` commands stream `bash_execution_update` chunks correlated by request `id` (Pi 0.82.0); include an `id` on the command to associate them.
+- `compact` responses include summary `usage`; `get_tokens`/`get_status` totals cover assistant messages, tool-reported usage, and summaries (Pi 0.81.0).
 - `@earendil-works/pi-coding-agent/rpc-entry` (Pi 0.80.3) launches Pi directly in RPC mode from an importing process.
 - RPC extension UI request/response types are exported from the public API (Pi 0.79.0).
 - Package asset path helpers are exported from the public API (Pi 0.79.0).
@@ -227,6 +280,20 @@ session.subscribe((event) => {
 - Add custom skills or prompt templates without touching disk
 - Share an event bus between the host app and loaded extensions
 
+To name an inline factory in the startup Extensions list (instead of `<inline:1>`), wrap it in an `InlineExtension` (Pi 0.80.4):
+
+```typescript
+import type { InlineExtension } from "@earendil-works/pi-coding-agent";
+
+const myProvider: InlineExtension = {
+  name: "my-provider",
+  factory: (pi) => { /* ... */ },
+};
+const loader = new DefaultResourceLoader({ extensionFactories: [myProvider] });
+```
+
+Bare factory functions are still accepted.
+
 ## Standalone Custom Tools
 
 Use `defineTool()` when you want a reusable custom tool definition outside `pi.registerTool(...)`:

package/skills/pi/references/settings.md

@@ -17,6 +17,7 @@ Edit JSON directly or use `/settings` for common interactive options.
   "defaultModel": "claude-sonnet-4-20250514",
   "defaultThinkingLevel": "medium",
   "hideThinkingBlock": false,
+  "showCacheMissNotices": false,
   "thinkingBudgets": {
     "minimal": 1024,
     "low": 4096,
@@ -26,8 +27,9 @@ Edit JSON directly or use `/settings` for common interactive options.
 }
 ```
 
-- Thinking levels: `"off"`, `"minimal"`, `"low"`, `"medium"`, `"high"`, `"xhigh"`
+- Thinking levels: `"off"`, `"minimal"`, `"low"`, `"medium"`, `"high"`, `"xhigh"`, `"max"` — `max` (Pi 0.80.6) is an opt-in level above `xhigh`, exposed only when the selected model supports it (some models, e.g. GPT-5.6, expose both). Custom themes can define `thinkingMax`; existing themes fall back to `thinkingXhigh`.
 - `hideThinkingBlock` hides visible thinking output in the UI
+- `showCacheMissNotices` (Pi 0.80.4) shows transcript notices for significant prompt-cache misses
 
 ## UI & Display
 
@@ -49,10 +51,28 @@ Edit JSON directly or use `/settings` for common interactive options.
 - `doubleEscapeAction`: `"tree"`, `"fork"`, or `"none"`
 - `treeFilterMode`: `"default"`, `"no-tools"`, `"user-only"`, `"labeled-only"`, or `"all"`
 - In `/tree`, `Shift+T` toggles timestamps on entry labels
-- `theme` also accepts `"light-name/dark-name"` (Pi 0.79.7) to switch themes automatically with the terminal color scheme; `/` is reserved in theme names for this. On first run Pi detects the terminal background and defaults to `dark` or `light`.
-- `outputPad` (Pi 0.80.3, `0` or `1`) sets horizontal padding for user messages, assistant messages, and thinking blocks.
+- `theme` also accepts `"light-name/dark-name"` (Pi 0.79.7) to switch themes automatically with the terminal color scheme; `/` is reserved in theme names for this. On first run Pi detects the terminal background and defaults to `dark` or `light`. `--use-theme <name[/name]>` (Pi 0.84.2) sets a per-run initial theme without changing saved settings.
+- `outputPad` (Pi 0.80.3, `0` or `1`) sets horizontal padding for user messages, assistant messages, and thinking blocks; custom message renderers receive it as `options.outputPad` (Pi 0.82.1).
 - `externalEditor` (Pi 0.80.3) sets the `Ctrl+G` external editor command and takes precedence over `$VISUAL`/`$EDITOR`.
 
+## Fullscreen TUI Mode
+
+Experimental since Pi 0.84.0:
+
+```json
+{
+  "tuiMode": "regular",
+  "fullscreenExitOutput": "transcript",
+  "fullscreenScrollbar": "auto"
+}
+```
+
+- `tuiMode`: `"regular"` (default) or `"fullscreen"`; `/settings` switches at runtime, `--tui-mode` overrides per run
+- `fullscreenExitOutput` (Pi 0.84.2): `"transcript"` prints the final transcript on exit; `"resume-hint"` restores the previous screen and prints only a resume hint
+- `fullscreenScrollbar`: `"auto"` shows while scrolling, `"always"` reserves the rightmost column, `"hidden"`
+- Fullscreen keeps editor/status/widgets/footer docked while the transcript scrolls independently; transcript search (`Ctrl+Shift+F`, Pi 0.84.2), page/half-page/line scrolling, and marked-message navigation are configurable via `tui.altScreen.*` keybindings
+- Theme tokens `scrollbarThumb`, `searchMatchBg`, `searchMatchText` are optional additions with fallbacks (`selectedBg`, `selectedBg`, `text`)
+
 ## Compaction, Branch Summary, Retry
 
 ```json
@@ -125,6 +145,7 @@ Notes:
 }
 ```
 
+- `shellPath` supports a leading `~` for the home directory (Pi 0.80.6).
 - `npmCommand` is argv-style and is used for npm lookup/install operations, including git-package installs. User-scoped npm packages install under `~/.pi/agent/npm/`; project-scoped npm packages install under `.pi/npm/`.
 - `terminal.showTerminalProgress` (default `false` since Pi 0.70.0) toggles OSC 9;4 progress reporting in supporting terminals (iTerm2, WezTerm, Windows Terminal, Kitty).
 - `terminal.imageWidthCells` (Pi 0.68.1) caps inline tool-output image width in terminal cells.
@@ -146,14 +167,19 @@ Notes:
 {
   "sessionDir": ".pi/sessions",
   "enabledModels": ["claude-*", "gpt-4o", "gemini-2*"],
+  "defaultTools": ["bash", "read", "edit"],
   "markdown": {
-    "codeBlockIndent": " "
+    "codeBlockIndent": "  ",
+    "mermaid": "streaming"
   }
 }
 ```
 
 When multiple sources specify a session directory, `--session-dir` takes precedence over `sessionDir` in settings. Pi 0.65.0 removed the old `session_directory` extension/settings hook.
 
+- `defaultTools` (Pi 0.84.2) picks the built-in tools enabled at startup (global or per project; project replaces global). An empty array starts with no built-in tools while keeping extension/SDK custom tools. `--tools` remains a strict allowlist for all tools, `--exclude-tools` filters the resulting list.
+- `markdown.mermaid` (Pi 0.84.0): `"off"`, `"final"`, or `"streaming"` — themed Unicode rendering of supported Mermaid diagrams in interactive messages. Pi 0.84.0 also renders terminal-friendly Unicode for LaTeX expressions.
+
 ## Project Trust
 
 Since Pi 0.79.0, project-local settings, resources, and packages (`.pi/` settings/extensions/skills/prompts/themes, project packages) only load after the project is trusted. `AGENTS.md` / `CLAUDE.md` context files load regardless of trust. Trust is a loading gate, not a sandbox.
@@ -236,7 +262,7 @@ pi [options] [@files...] [messages...]
 --provider <name>                e.g. anthropic, openai, google
 --model <pattern>                supports provider/id and optional :<thinking>
 --api-key <key>
---thinking off|minimal|low|medium|high|xhigh
+--thinking off|minimal|low|medium|high|xhigh|max
 --models <patterns>              Ctrl+P cycling allowlist
 --list-models [search]
 
@@ -270,6 +296,8 @@ pi [options] [@files...] [messages...]
 # prompts and misc
 --system-prompt <text>           replace default
 --append-system-prompt <text>    repeatable; appended with double newlines
+--tui-mode regular|fullscreen    experimental fullscreen TUI (Pi 0.84.0)
+--use-theme <name[/name]>        per-run initial theme, saved settings untouched (Pi 0.84.2)
 --verbose
 ```
 
@@ -281,14 +309,20 @@ pi [options] [@files...] [messages...]
 | `PI_CODING_AGENT_SESSION_DIR` | Override session storage; `--session-dir` still wins |
 | `PI_PACKAGE_DIR` | Override package storage directory |
 | `PI_SKIP_VERSION_CHECK` | Skip the `pi.dev` latest-version request at startup |
-| `PI_OFFLINE` | Disable all startup network operations (update checks, telemetry) |
+| `PI_OFFLINE` | Disable all startup network operations (update checks, telemetry, model catalog refresh) |
 | `PI_TELEMETRY` | Override install/update telemetry and provider attribution headers (`1`/`0`); does not gate update checks |
 | `PI_CACHE_RETENTION` | Set to `long` for extended prompt-cache retention where supported |
 | `PI_OAUTH_CALLBACK_HOST` | Bind the OAuth callback server to a custom interface |
-| `PI_CODING_AGENT=true` | Set automatically at startup so subprocesses can detect Pi |
+| `PI_SHARE_VIEWER_URL` | Override the base URL used by `/share` |
+| `PI_HARDWARE_CURSOR` | `1` shows the hardware cursor (see terminal-setup docs) |
+| `PI_TUI_ESC_TIMEOUT` | Ms to wait after a lone ESC before treating it as Escape (default 100 over SSH, 10 otherwise; Pi 0.84.2) — raise it if Alt-key input is misread |
+| `PI_EXPERIMENTAL` | `1` enables experimental features, e.g. strict JSON-schema sampling for built-in tools (Pi 0.84.2) |
+| `AI_AGENT=pi` / `PI_CODING_AGENT=true` | Set automatically by CLI/RPC entry points so child processes can detect Pi (generic vs Pi-specific marker) |
 | `VISUAL`, `EDITOR` | External editor for `Ctrl+G` (the `externalEditor` setting takes precedence) |
 | Provider API key env vars | See `references/providers.md` |
 
+Bash-tool commands additionally receive session state since Pi 0.82.0: `PI_SESSION_ID`, `PI_SESSION_FILE`, `PI_PROVIDER`, `PI_MODEL`, and `PI_REASONING_LEVEL` (resolved per command; not injected into user-entered `!`/`!!` commands). When asked which model is running, inspect these instead of inferring from the system prompt.
+
 ## Practical Notes
 
 - Project settings are good for shared package/resource configuration.