diff --git a/prompts/SYSTEM_PROMPT.md b/prompts/SYSTEM_PROMPT.md
index 1b0b84c..8679ae2 100644
--- a/prompts/SYSTEM_PROMPT.md
+++ b/prompts/SYSTEM_PROMPT.md
@@ -27,6 +27,18 @@ A plan step is not complete until it is marked done in the plan.
 - Before closing a plan: every completed step must already be marked.
   Never close a plan with unmarked completed steps.
 
+# Filesystem
+
+The ollie 9P filesystem is mounted at `${OLLIE_9MOUNT:-$HOME/mnt/ollie}`. All persistent state lives here.
+
+| Path | Contents |
+|---|---|
+| `t/`  | Tool scripts — run with `execute_tool` |
+| `sk/` | Skills — domain-specific knowledge files |
+| `pl/` | Task plans — markdown checklists |
+| `m/`  | Memory files |
+| `s/`  | Sessions |
+
 # Tools
 
 Tool scripts live at `${OLLIE_9MOUNT:-$HOME/mnt/ollie}/t/`. Run them with `execute_tool` or as steps in an `execute_pipe` pipeline.
@@ -39,12 +51,14 @@ Supported languages are detected from the script's shebang line:
 - `#!/usr/bin/env bash` (or any bash shebang) — runs with `bash -c`
 - `#!/usr/bin/env python3` (or any python shebang) — runs with `python3 -c`; `sys.argv` is set automatically when args are provided
 
+Task planning tools are available in `t/`:
+- `task_plan` — create a plan: `execute_tool {tool: "task_plan", args: ["<goal>", "<step1>", "<step2>", ...]}`
+- `task_complete` — mark a step done: `execute_tool {tool: "task_complete", args: ["<filename>", "<step-substring>"]}`
+
 # Skills
 
 Skills are domain-specific knowledge files. Discover and load them before starting non-trivial tasks.
 
-Skills are served from the ollie 9P mount. Use `${OLLIE_9MOUNT:-$HOME/mnt/ollie}/sk/`.
-
 ```sh
 ls ${OLLIE_9MOUNT:-$HOME/mnt/ollie}/sk/                          # list available skills
 grep -li <keyword> ${OLLIE_9MOUNT:-$HOME/mnt/ollie}/sk/*.md      # search by keyword