diff --git a/README.md b/README.md index 04b4779..93ccd0b 100644 --- a/README.md +++ b/README.md @@ -53,15 +53,15 @@ Everything lives under `$XDG_CONFIG_HOME/ollie/` (default: `~/.config/ollie/`) | Capability | How | |---|---| -| **CLI** | `o` — unified namespace CLI: `read`, `write`, `ls`, `ctl`, `prompt`, `tui`, `env`, `generate` | +| **CLI** | `o` — unified namespace CLI: `read`, `write`, `ls`, `new`, `ctl`, `prompt`, `tui`, `env`, `generate` | | **Terminal TUI** | `o tui` — tmux layout, no widgets | | **One-shot LLM** | `o generate "explain monads"` or `echo "prompt" \| o generate` | -| **Run an agent** | Create a session + agent via 9P — then connect with any frontend ([ellie](doc/ellie.md), KDE (GUI/KRunner/Kate), `o tui`) | +| **Run an agent** | Create a session + agent via 9P — then connect with any frontend ([ellie](doc/ellie.md), KDE (GUI/KRunner/Kate), acme, `o tui`) | | **Remote execution** | Set `remote=user@host` in session config | | **Background processes** | Pass `"background": true` to any tool — runs with no timeout, output pushed to agent prompt on completion | | **Parallel execution** | Non-conflicting tool calls within a turn run in parallel (scope-based conflict scheduling) | -| **Agents** | Write a JSON config in `data/agents/` with agent prompt in `data/prompts/` | -| **Tools** | Drop an executable + `.meta` in `~/.config/ollie/tools/` — load at runtime via `/tool_load` | +| **Agents** | Write a JSON config in `~/.config/ollie/agents/` with agent prompt in `~/.config/ollie/prompts/` | +| **Tools** | Drop an executable + `.meta` in `~/.config/ollie/tools/` — load at runtime via `tool_load` | | **Domain skills** | Load markdown skill modules at runtime | ## Repository layout @@ -70,28 +70,42 @@ Single Go module with one Git submodule (kde). | Directory | Language | Description | |-----------|----------|-------------| -| `agent/` | Go | Agent loop, history, prompt resolution, context compaction | -| `backend/` | Go | LLM providers (Anthropic, OpenAI, Ollama, Gemini, Copilot, CodeWhisperer) | -| `toolsrv/` | Go | Tool server client library (9P client for toolsrv) | -| `cmd/toolsrv/` | Go | Tool execution server (separate 9P server, sandboxed execution) | -| `tools/lsp/` | Go | LSP bridge daemon + client library (gopls, clangd, intelephense) | -| `session/` | Go | Session lifecycle, config, persistence | -| `fs/` | Go | 9P filesystem tree (EDSL-declared namespace, ctl dispatch) | -| `fsedsl/` | Go | Filesystem declaration EDSL (generic, reusable) | -| `bypass/` | Go | Bypass broker (sandbox escape approval) | -| `sandbox/` | Go | Landlock sandbox config | -| `lib9p/` | Go | 9P protocol library + native C client | | `cmd/olliesrv/` | Go | The main server binary (session/agent 9P namespace) | -| `cmd/toolsrv/` | Go | Tool execution server (spawned by olliesrv, separate 9P server) | +| `cmd/toolsrv/` | Go | Tool execution server (separate 9P server, sandboxed execution) | | `cmd/ollie-9p/` | Go | 9P client CLI | -| `cmd/ollie-remote/` | Go | Remote execution binary | +| `toolsrv/` | Go | Tool server client library + tool discovery (9P client for toolsrv) | +| `virtfs/` | Go | Filesystem declaration EDSL (generic, reusable) | +| `lib9p/` | Go | 9P protocol library + native C client | +| `tools/lsp/` | Go | LSP bridge daemon + client library (gopls, clangd, intelephense) | +| `env/` | Go | Environment variable defaults | +| `format/` | Go | Output formatting | +| `log/` | Go | Structured logging | +| `paths/` | Go | XDG path resolution | | `kde/` | C++/Qt6 | KDE plasmoid, GUI, Kate plugin, KRunner *(submodule)* | -| `data/agents/` | JSON | Agent configs | +| `contrib/elisp/` | Emacs Lisp | ellie.el — Emacs frontend | +| `data/agents/` | JSON | Agent configs (installed to `~/.config/ollie/agents/`) | | `data/prompts/` | Markdown | System prompt templates | | `data/tools/` | Mixed | Tool executables (scripts + compiled) + `.meta` sidecar files | | `data/skills/` | Markdown | Domain knowledge modules | +| `data/scripts/` | Shell | CLI scripts (`o`, `ollie-remount`, acme helpers) | | `doc/` | Markdown | Architecture docs, usage guide | +### Internal packages (not importable outside their command) + +| Package | Parent | Description | +|---------|--------|-------------| +| `agent` | `cmd/olliesrv` | Agent loop, history, prompt resolution, context compaction | +| `backend` | `cmd/olliesrv` | LLM providers (Anthropic, OpenAI/OpenRouter, Ollama, Gemini, Copilot, Kiro) | +| `session` | `cmd/olliesrv` | Session lifecycle, config, persistence | +| `fs` | `cmd/olliesrv` | 9P filesystem tree (virtfs-declared namespace, ctl dispatch) | +| `bypass` | `cmd/olliesrv` | Bypass broker (sandbox escape approval) | +| `toolclient` | `cmd/olliesrv` | Spawns and connects to toolsrv | +| `prompts` | `cmd/olliesrv` | Prompt template resolution | +| `registry` | `cmd/toolsrv` | Per-agent tool registry | +| `exec` | `cmd/toolsrv` | Sandboxed tool execution | +| `sandbox` | `cmd/toolsrv` | Landlock sandbox config and wrapper | +| `server` | `cmd/toolsrv` | 9P filesystem + proc management | + ## Architecture ```mermaid @@ -126,6 +140,7 @@ graph TB ANT[Anthropic] COP[Copilot] KIRO[Kiro] + GEM[Gemini] end TUI & SCRIPTS & ACME --> O @@ -135,7 +150,7 @@ graph TB SESS --> AG SESS -->|9P over Unix socket| TOOLS - AG -->|streaming| OLL & OAI & ANT & COP & KIRO + AG -->|streaming| OLL & OAI & ANT & COP & KIRO & GEM TOOLS --> EXEC ``` @@ -156,4 +171,4 @@ filesystem, see [`doc/9p.md`](doc/9p.md). ## License -GPLv3 \ No newline at end of file +GPLv3 diff --git a/doc/usage.md b/doc/usage.md index 3521df4..4a3d804 100644 --- a/doc/usage.md +++ b/doc/usage.md @@ -32,20 +32,21 @@ Valid levels: `debug`, `info`, `warn`, `error`. ## `o` — Terminal CLI `o` is a multi-call shell script that wraps the 9P namespace into a human-friendly interface. -Installed to `~/bin/o` by `just install-data`. +Installed to `~/.local/bin/o` by `just install-data`. ### Quick start ```sh -eval $(o env myproject default) # set session + agent context -o prompt # interactive prompt REPL +o new myproject # create a session +o myproject new default # create an agent in that session +o myproject/default tui # launch the tmux TUI ``` -In another terminal: +Or without the TUI: ```sh -eval $(o env myproject default) -o read chat # stream chat output +o myproject/default prompt # interactive prompt REPL (one terminal) +o myproject/default read chat # stream chat output (another terminal) ``` @@ -61,59 +62,70 @@ o read chat # stream chat output
+### Context model + +Context is determined by the first argument — not by environment variables: + +``` +o Root level (no context) +o Session level +o / Agent level +``` + +The slash in `/` disambiguates session-only from session+agent. +If no command follows the context, `o` defaults to `ls`. + ### Commands -| Command | Description | -|---|---| -| `o prompt` | Interactive multi-line prompt REPL | -| `o read [-l] ` | Read a file (-l: loop, re-read after each return) | -| `o write [data]` | Write to any file | -| `o ls [path]` | List namespace entries | -| `o ctl [args]` | Send control command (rdwr: gets response) | -| `o generate ` | One-shot generation (no session needed) | -| `o tui` | Launch tmux TUI | -| `o env [session] [agent]` | Print export statements for context | - -### Context - -Context is set via `$session` and `$agent` environment variables: - -```sh -eval $(o env myproject default) # export session=myproject agent=default - -# Now every o command addresses the right agent automatically: -o ctl model # → shows current model -o read -l statewait -``` +| Command | Context | Description | +|---|---|---| +| `o ls` | root | List sessions, models, backends | +| `o new [cwd]` | root | Create a new session | +| `o generate ` | root | One-shot generation (no session needed) | +| `o ls` | session | List session contents | +| `o new [cwd]` | session | Create a new agent | +| `o ctl ` | session | Session control | +| `o / prompt` | agent | Interactive multi-line prompt REPL | +| `o / tui` | agent | Launch tmux TUI | +| `o / ctl ` | agent | Agent control (stop, compact, model, etc.) | +| `o / log [-n N]` | agent | Show last N lines of agent log | +| `o read [-l] ` | any | Read a file (-l: loop) | +| `o write [data]` | any | Write to a file (args or stdin) | +| `o env` | any | Print export statements for current context | +| `o help` | any | Show help | ### Path resolution -Paths are slash-delimited 9P namespace paths. The `o` CLI resolves them based -on context: +When you supply a context, `o` sets a **prefix**: +- Session: `session/` +- Agent: `session//agent/` -| Path type | Example | Resolves to | -|---|---|---| -| Root-level | `o read models` | `models` (no context needed) | -| Agent file | `o read state` | `session/$session/agent/$agent/state` | -| Session file | `o read env` | `session/$session/env` | -| Full path | `o read session/foo/agent/bar/chat` | Used as-is | +Paths given to `read`, `write`, and `ls` are resolved relative to this prefix. +A leading `/` escapes the prefix and addresses the root namespace directly: -Root-level paths (`session/`, `agents`, `models`, `help`, `ctl`, `tools`) and -absolute paths (starting with `session/`) bypass context resolution entirely. -Agent-level files (`chat`, `state`, `prompt`, `statewait`, `cfg`, `plan`, -`log`, `cost`, `offset`, `context`, `systemprompt`) require both `$session` -and `$agent`. Session-level files (`env`, `ctl`, `id`, `name`, `paused`) -require only `$session`. +| Invocation | Resolves to | +|---|---| +| `o myproj/coding read chat` | `session/myproj/agent/coding/chat` | +| `o myproj/coding read cfg` | `session/myproj/agent/coding/cfg` | +| `o myproj read env` | `session/myproj/env` | +| `o myproj/coding read /models` | `models` (root) | +| `o myproj/coding ls /bypass` | `bypass` (root) | +| `o read models` | `models` (no context, no prefix) | + +The leading slash is necessary when you want to access root-level paths +(like `models`, `backends`, `bypass/`) while in session or agent context. +Without it, the path is concatenated with the context prefix. ### Prompt REPL -`o prompt` is an interactive multi-line REPL with readline editing: +`o / prompt` is an interactive multi-line REPL: ``` -$ o prompt -[myproject/default] +$ o myproj/coding prompt +[myproj/coding] Type . on a blank line to send, Ctrl+D to exit. -Slash commands: /stop /compact /kill /clear /model /backend /cwd /help +/ prefix runs ctl: /stop, /compact, /model X, /inject X +!q exits the TUI. > write a fibonacci function in python and include a test case @@ -127,13 +139,16 @@ Slash commands control the agent without sending a prompt: |---|---| | `/stop` | Interrupt running agent | | `/compact` | Compact context window | -| `/kill` | Kill the session (exits REPL) | -| `/clear` | Clear history | | `/model X` | Switch model | -| `/backend X` | Switch backend | | `/cwd X` | Change working directory | -| `/quit` or `/q` | Kill the tmux TUI | -| `/help` | List commands | +| `!q` | Kill the tmux TUI session and exit | + +Piped input is also supported: + +```sh +echo "explain this error" | o myproj/coding prompt +cat file.go | o myproj/coding write prompt +``` ### One-shot generation @@ -145,18 +160,13 @@ cat error.log | o generate # pipe content as the prompt echo "write a haiku about filesystems" | o generate | wc -w ``` -Chain multiple steps: - -```sh -echo "list 5 blog post ideas" | o generate | head -3 | o generate -``` - --- ## `o tui` — tmux Terminal UI -`o tui` composes `o read chat` and `o prompt` into a single tmux layout — two shell -commands in two panes, with agent state in the tmux status bar: +`o / tui` composes `o read chat` and `o prompt` into a single +tmux layout — two shell commands in two panes, with agent state in the tmux +status bar:

Screenshot of the o tui tmux layout: chat pane on top, prompt REPL on bottom, agent state in status bar @@ -165,17 +175,15 @@ commands in two panes, with agent state in the tmux status bar: ### Usage ```sh -eval $(o env myproject default) -o tui +o myproj/coding tui ``` -1. Requires `$session` and `$agent` to be set. -2. Creates a tmux session named `o-`. -3. Two panes: - - **Pane 0** (top, 80%): `o read chat` — streams filtered chat output (markers stripped server-side). +1. Creates a tmux session named `o-`. +2. Two panes: + - **Pane 0** (top, 80%): streams chat output (log tail + live `read chat`). - **Pane 1** (bottom, 20%): `o prompt` — multi-line REPL (`/` prefix runs ctl commands). -4. A background loop reads `statewait` and updates the tmux status bar instantly. -5. Focuses the prompt pane and attaches. +3. A background loop reads `statewait` and updates the tmux status bar on each state transition. +4. Focuses the prompt pane and attaches. No polling. Each read blocks until data arrives. The status bar updates on every state transition via `statewait` — zero CPU when idle. @@ -189,10 +197,28 @@ state transition via `statewait` — zero CPU when idle. --- +## `o env` — Context export + +`o env` prints export statements for the current context. This is useful for +other tools (raw `ollie-9p` commands, scripts) that need to know which session +and agent are being addressed: + +```sh +eval $(o myproj/coding env) +# exports OLLIE_SESSION=myproj OLLIE_AGENT=coding +ollie-9p read session/$OLLIE_SESSION/agent/$OLLIE_AGENT/stats +``` + +Note: `o` itself does not read these environment variables — it determines +context purely from its positional arguments. + +--- + ## Other front-ends - **[ellie](ellie.md)** — Emacs session tree, chat, multi-agent -- **KDE** — plasmoid, standalone GUI, Kate plugin, KRunner +- **[KDE](kde-gui.md)** — plasmoid, standalone GUI, Kate plugin, KRunner +- **acme** — Plan 9 editor (scripts in `data/scripts/acme/`) -For technical details (generation parameters, sandboxing, multi-agent workflows, -tool discovery, and the raw 9P filesystem), see [doc/whitepaper.md](whitepaper.md). \ No newline at end of file +For technical details on the raw 9P filesystem, see [`doc/9p.md`](9p.md). +For tool authoring, see [`doc/writing-tools.md`](writing-tools.md).