@cmodjs/core 1.6.0 → 2.0.1

package/docs/commands.md

@@ -12,7 +12,7 @@ Each command below restates what `cmod <command> --help` prints, for a mod autho
 cmod new <name> [--project]
 ```
 
-Creates a mod in `./<name>`: a `defineMod` with one hook, one pane, and one render of the band above the prompt, a Skill, a test, and a `CHANGELOG.md` ([mod.md](mod.md#release-notes)). Then installs its packages.
+Creates a mod in `./<name>`: a `defineMod` with one listener, one pane, and one render of the band above the prompt, a Skill, a test, and a `CHANGELOG.md` ([mod.md](mod.md#release-notes)). Then installs its packages.
 
 - `<name>` is lowercase letters, digits, and `-`, starting with a letter.
 - `--project` creates a project-scope plugin in `./.claude/skills/<name>/` of the repository in the current folder, which loads for everyone who trusts the repository. Run it in the repository's root.
@@ -98,7 +98,7 @@ Releases the mod at `path` (default: the current folder) at the version in its `
 4. Commits the release as the `release` branch, and builds the release archive from that commit.
 5. Builds the program `cli/` declares, and writes `SHA256SUMS` for every file of the release.
 6. Writes `.claude-plugin/marketplace.json`, listing the archive and the cmod plugin. The cmod entry points at the `release` branch of `heyJordanParker/cmod`, so the mod's `"cmod"` dependency resolves in this marketplace and follows each cmod release with no new publish of the mod.
-7. Commits that file, tags `v<version>`, pushes the tag and the `release` branch, creates the GitHub release with every file, then pushes the branch that holds `marketplace.json`. So Claude Code never updates anyone to a release whose files are not up yet. Its notes are the version's section of the committed `CHANGELOG.md`, above GitHub's compare link ([mod.md](mod.md#release-notes)). `--dry-run` prints them.
+7. Commits that file, tags `v<version>`, tags the release commit `<name>--v<version>`, which Claude Code installs a version range from when a marketplace lists the mod from its git repository, pushes both tags and the `release` branch, creates the GitHub release with every file, then pushes the branch that holds `marketplace.json`. So Claude Code never updates anyone to a release whose files are not up yet. Its notes are the version's section of the committed `CHANGELOG.md`, above GitHub's compare link ([mod.md](mod.md#release-notes)). `--dry-run` prints them.
 8. Prints the link to paste as the Repository when you submit the mod at [claude.ai/directory/manage](https://claude.ai/directory/manage), such as `https://github.com/owner/greeter/tree/release`. The portal reads the branch from the link, so the directory follows `release`.
 
 The release keeps `package.json`, so Claude Code still installs the mod's packages for its install and uninstall steps. It keeps `bun.lock` too, unless a `"files"` list leaves it out. `main` keeps the source only.
@@ -114,7 +114,7 @@ It needs, and refuses to start without:
 - a `package.json` with no dependency on a `file:` or `link:` path, which no user has
 - a `"files"` list, when `package.json` has one, whose every path matches a committed file
 - a `"program"` in the `cmod` key of `package.json` that names the program `cli/` declares, when `cli/` declares one
-- for a real publish, a working tree with no uncommitted changes, a `v<version>` tag that does not exist yet, a `git push` to `origin` that succeeds, and the `gh` command, which creates the GitHub release
+- for a real publish, a working tree with no uncommitted changes, no `v<version>` or `<name>--v<version>` tag yet, a `git push` to `origin` that succeeds, and the `gh` command, which creates the GitHub release
 
 `--dry-run` builds everything into a temporary folder, and pushes nothing. It writes `marketplace.json` there too, so the working tree stays as it was. It allows uncommitted changes and an existing tag, and builds from the last commit. It prints the folder that holds the release, so you can read what the directory will read.
 
@@ -179,7 +179,7 @@ Lists every plugin Claude Code has installed or linked, with its version. A mod
 
 ```text
 cmod history sessions [--all] [--since 2d] [--running] [--project <folder>] [--json]
-cmod history search <text> [--all] [--since 2d] [--running] [--project <folder>] [--session <id> [--task <id>] [--turn <id>] [--context <n>] [--goal]] [--limit <n>] [--before <session>:<id>] [--json]
+cmod history search <text> [--all] [--since 2d] [--running] [--project <folder>] [--session <id> [--task <id>] [--turn <id>] [--context <n>] [--agent]] [--limit <n>] [--before <session>:<id>] [--json]
 cmod history messages [<session>] [--task <id>] [--turn <id>] [--after <id>] [--last <n>] [--json]
 cmod history turns [<session>] [--task <id>] [--context <n>] [--last <n>] [--json]
 cmod history contexts [<session>] [--task <id>] [--json]
@@ -194,7 +194,7 @@ cmod history index
 Reads every conversation Claude Code saved, as the sessions, contexts, turns, tasks, and messages `sessions` answers a mod ([session.md](session.md)), so a person or an agent reads them in a terminal without parsing a transcript.
 
 - `sessions` lists the conversations of the current folder's project, or `--project`'s, newest first, each with its time, id, and title. `--all` lists every project's, `--since` keeps those active in the last `30m`, `6h`, `2d`, or `1w`, or since a date such as `2026-10-01`, and `--running` keeps those open in Claude Code now.
-- `search` prints the newest 50 messages whose text or tool input contains the text, at least 3 characters, any case, each with its time and session. Each tool input value is searched in its first 1,024 characters, and tool results are not searched. `--all`, `--since`, and `--running` scope it as they scope `sessions`, `--limit` changes the 50, and `--before <session>:<id>` prints the page after the message it names, the last one of the page before. `--session` searches one conversation, and inside it `--task` searches a task and the tasks it started, `--turn` one turn, `--context` one context, and `--goal` the tasks of the conversation or of `--task`.
+- `search` prints the newest 50 messages whose text or tool input contains the text, at least 3 characters, any case, each with its time and session. Each tool input value is searched in its first 1,024 characters, and tool results are not searched. `--all`, `--since`, and `--running` scope it as they scope `sessions`, `--limit` changes the 50, and `--before <session>:<id>` prints the page after the message it names, the last one of the page before. `--session` searches one conversation, and inside it `--task` searches a task and the tasks it started, `--turn` one turn, `--context` one context, and `--agent` keeps the main agent's messages, or with `--task` that task's own, without the tasks it started.
 - `messages` prints one conversation's messages in order: the session named, or the one the command runs in when Claude Code runs it. `--task` reads a task's messages instead, `--turn` keeps one turn's, `--after` starts after a message id, and `--last` keeps that many from the end.
 - `turns` prints the conversation's turns, each opened by a prompt and closed when Claude finishes. `--context` keeps one context's, and `--last` keeps that many from the end.
 - `contexts` prints the conversation's contexts, a new one after each compaction.
@@ -205,7 +205,7 @@ Reads every conversation Claude Code saved, as the sessions, contexts, turns, ta
 - `state` prints what a running conversation resumes from.
 - `follow` prints each message, turn, context, task, and goal change of the sessions named as Claude Code saves it, until stopped with Ctrl+C. `--running` follows every running session of the project too, and each one that starts later. The cmod plugin runs it to hear other running sessions.
 - `--json` prints JSON instead of one line each.
-- `index` reads every saved conversation into the index at `~/.local/share/cmod/history-5.sqlite`, which holds where each message is in its transcript, not a copy of it, and each conversation's turns, contexts, and tasks. While a mod other than cmod holds the `conversation` permission, the cmod plugin runs it in the background at each session's start, and a second run while one is running exits at once. The index forgets a conversation whose transcript is deleted.
+- `index` reads every saved conversation into the index at `~/.local/share/cmod/history-6.sqlite`, which holds where each message is in its transcript, not a copy of it, and each conversation's turns, contexts, and tasks. While a mod other than cmod holds the `conversation` permission, the cmod plugin runs it in the background at each session's start, and a second run while one is running exits at once. The index forgets a conversation whose transcript is deleted.
 
 ## Run in an install script
 
@@ -259,7 +259,7 @@ Changes Claude Code settings for a set-up mod, as its `settings.user.update`, `s
 cmod rules [--project <folder>] [--json]
 ```
 
-Links the `rules/` folder of each mod that is on and has the `prompt` permission into Claude Code's rules folder, and removes the link of each that is off, lacks the permission, or is removed ([mod.md](mod.md#rules)). A mod the person installed links at `<config>/rules/<mod>`, and one turned on only for `--project`, the current folder when it is left out, links at `<project>/.claude/rules/<mod>`. It prints the rules it linked. With `--json` it prints `{ linked, placed, removed }`, where `placed` and `removed` name the mods whose link this run made or took away. `cmod install`, `update`, `remove`, `enable`, `disable`, `permission`, `setup`, `teardown`, `link`, and `unlink` run it after they succeed, and the cmod plugin runs it at each session's start and each prompt.
+Links the `rules/` folder of each mod that is on and has the `prompt` permission into Claude Code's rules folder, and removes the link of each that is off, lacks the permission, or is removed ([mod.md](mod.md#rules)). A mod the person installed links at `<config>/rules/<mod>`, and one turned on only for `--project`, the current folder when it is left out, links at `<project>/.claude/rules/<mod>`. It prints the rules it linked. With `--json` it prints `{ linked, placed, removed, watched }`, where `placed` and `removed` name the mods whose link this run made or took away, and `watched` lists the files and folders whose change can change the next run. `cmod install`, `update`, `remove`, `enable`, `disable`, `permission`, `setup`, `teardown`, `link`, and `unlink` run it after they succeed. The cmod plugin runs it when a session starts and before a prompt, each time only when a `watched` path changed since its last run.
 
 ### cmod option
 

package/docs/dependencies.md

@@ -58,7 +58,7 @@ A provider that changes or removes a method breaks every mod that calls it, and
 "dependencies": [{ "name": "cmod", "version": "^1.0.0" }, { "name": "tracer", "version": "^1.0" }]
 ```
 
-Claude Code checks the range against the `version` in the provider's `plugin.json` when it loads your mod ([Claude Code's plugin dependencies](https://code.claude.com/docs/en/plugins/dependencies)). cmod is a provider too: a mod built on `@cmodjs/core` 1.0 calls methods only the 1.0 cmod plugin has, so `cmod new` writes the `^1.0.0` range.
+Claude Code checks the range against the provider's `version` when it loads your mod. Where a marketplace lists the provider from its git repository, as every mod's marketplace lists cmod, Claude Code also installs it at the highest `<name>--v<version>` tag inside the range, which `cmod publish` pushes for every release ([Claude Code's plugin dependencies](https://code.claude.com/docs/en/plugins/dependencies)). cmod is a provider too: a mod built on `@cmodjs/core` 1.0 calls methods only the 1.0 cmod plugin has, so `cmod new` writes the `^1.0.0` range.
 
 Claude Code then lays tracer's types file beside the mod's own types, so the call is typed:
 

package/docs/index.md

@@ -16,7 +16,7 @@ my-mod/
 ├── .oxlintrc.json                      leaves .claude-plugin/types/ out of the lint
 ├── hooks/hooks.json                    names hooks/register.ts as the hooks module
 ├── hooks/register.ts                   registerMod(addHook, myMod, options), and nothing else
-├── src/mod.tsx                         defineMod: the state, the setup, the hooks, the panes
+├── src/mod.tsx                         defineMod: the state, the setup, the listeners, the panes
 ├── src/state.ts                        the starting state
 ├── src/panes/prompts.tsx               one definePane per file
 ├── src/components/prompt-count.tsx     what the panes and renders draw with
@@ -57,7 +57,7 @@ A program the mod needs lives as source in `cli/`. `cmod publish` builds it and
 | Question | File |
 | --- | --- |
 | How do I define a mod, and what can `mod` call? | [mod.md](mod.md) |
-| How does a mod watch something outside Claude Code, tell Claude about it, or ask a model? | [mod.md](mod.md#mod-every-and-mod-after), [session.md](session.md#send-to-a-conversation) |
+| How does a mod watch something outside Claude Code, tell Claude about it, or ask a model? | [mod.md](mod.md#mod-every-and-mod-after), [session.md](session.md#act-on-a-session) |
 | How does a mod give Claude a prompt on a schedule, like `/loop`? | [jobs.md](jobs.md#schedule) |
 | How do I keep values across prompts, sessions, and projects? | [state.md](state.md) |
 | How does a person set a token, a URL, or a setting for my mod, for themselves or their team? | [options.md](options.md) |
@@ -66,7 +66,8 @@ A program the mod needs lives as source in `cli/`. `cmod publish` builds it and
 | How does the person finish a setup step only they can do, such as signing in? | [install-steps.md](install-steps.md#steps-in-claude-code) |
 | How do I add my mod's own page to `/mods`, or change Claude Code's settings? | [mod.md](mod.md#mod-settings), [mod.md](mod.md#settings-claude-code-s-settings-files) |
 | How does my mod read and follow this conversation, its tasks, or earlier ones, search them, or start a task? | [session.md](session.md) |
-| How do I listen to a Claude Code hook, and what can a listener answer? | [hooks.md](hooks.md) |
+| How does my mod hear a prompt, a tool call, a permission dialog, or a model switch, and change or refuse it? | [session.md](session.md#listen-to-a-record), [session.md](session.md#the-session-s-own-events) |
+| How does my mod hear, change, or run any slash command, a built-in such as `/compact` included? | [mod.md](mod.md#commands-slash-commands) |
 | How do I draw a pane, change a row Claude Code draws, or ask the person something? | [ui.md](ui.md) |
 | How do I add a slash command, a tool, a permission rule, a check, a prompt, a schedule, a status line, or a background program? | [jobs.md](jobs.md) |
 | How does one mod call another? | [dependencies.md](dependencies.md) |
@@ -81,20 +82,20 @@ A mod imports from the files below, and every other file in this package is inte
 | Import | Exports | Docs |
 | --- | --- | --- |
 | `mod.js` | `defineMod`, `ModDefinition`, `Mod`, `Disposable`, `Permissions`, `Allowed`, `JobContext`, `Job`, `PaneHandle`, `longestMs`, `messageOf` | [mod.md](mod.md), [jobs.md](jobs.md) |
-| `mod.js` | `settings`, `SettingsFile`, `model`, `fs`, `Entry`, `Stat`, `FileMatch`, `metadata`, `Metadata`, `process`, `http`, `Imports` | [mod.md](mod.md) |
-| `mod.js` | `sessions`, `Sessions`, `Session`, `SessionScope`, `SessionSend`, `SessionEvent`, `SessionPayload`, `TaskStartArgs`, `Usage`, `Matcher` | [session.md](session.md) |
-| `mod.js` | `Message`, `MessageType`, `Prompt`, `Note`, `Reply`, `ToolCall`, `ToolResult`, `Skill`, `Document`, `Notice`, `Sender`, `SenderType`, `Person`, `Schedule`, `Channel`, `Sdk` | [session.md](session.md#message) |
-| `mod.js` | `Context`, `ContextStart`, `ContextEnd`, `Turn`, `TurnStart`, `TurnEnd`, `Goal`, `GoalChange`, `Task`, `TaskStart`, `TaskEnd` | [session.md](session.md#turn-context-goal-and-task) |
-| `mod.js` | `Hook`, `HookInput`, `HookListener`, `HookOptions`, `HookAnswer`, and one constant for each of Claude Code's 33 hooks | [hooks.md](hooks.md) |
+| `mod.js` | `settings`, `SettingsFile`, `SettingsFileChange`, `SettingsFileChangeListener`, `config`, `ConfigRow`, `commands`, `CommandInfo`, `CommandRun`, `CommandRunAnswer`, `CommandRunResult`, `model`, `fs`, `Entry`, `Stat`, `FileMatch`, `FileChange`, `metadata`, `Metadata`, `process`, `http`, `Imports` | [mod.md](mod.md) |
+| `mod.js` | `sessions`, `Sessions`, `Session`, `CurrentSession`, `TaskOwner`, `RecordPart`, `OnMethods`, `BeforeMethods`, `Listener`, `ListenerTime`, `ReadOptions`, `AnswerOptions`, `Refusal`, `SearchOptions`, `SessionScope`, `SessionSend`, `TaskStartArgs`, `ToolCallArgs`, `Usage`, `Surface`, `Matcher` | [session.md](session.md) |
+| `mod.js` | `Placement`, `ProjectSetup`, `SessionStartAsked`, `SessionStarted`, `SessionEnded`, `ModelChange`, `CwdChange`, `DirectoryAdd`, `Worktree`, `Notification`, `Elicitation`, `ElicitationResponse`, `McpServer`, `PermissionUpdate`, `PermissionCheck`, `PermissionRequest`, `PermissionDenial`, `BatchedCall`, `ToolBatch`, `Todo`, `ReplyDisplay` | [session.md](session.md#the-session-s-own-events) |
+| `mod.js` | `Message`, `MessageType`, `Prompt`, `Note`, `Reply`, `ToolCall`, `ToolResult`, `Skill`, `InstructionFile`, `Notice`, `Sender`, `SenderType`, `Person`, `Mod`, `Schedule`, `Channel`, `Sdk`, `Engine` | [session.md](session.md#message) |
+| `mod.js` | `Context`, `Turn`, `Goal`, `Task` | [session.md](session.md#turn-context-goal-and-task) |
 | `mod.js` | `Toast`, `ToastId`, `ToastType`, `ToastOptions` | [ui.md](ui.md#toasts) |
 | `mod.js` | `Place`, `PlaceProps`, `ToolGroup`, `ToolProgress`, `CommandOutput`, `AskUserQuestion`, `Spinner`, `TurnDuration`, `SessionMode`, `PromptHint`, `AbovePrompt` | [ui.md](ui.md#places-change-a-row-claude-code-draws) |
 | `mod.js` | `notInstalled` | [dependencies.md](dependencies.md) |
 | `installer.js` | `defineStep`, `Step` | [install-steps.md](install-steps.md#steps-in-claude-code) |
 | `register.js` | `registerMod`, `registerPermissionCheck`, `registeredMod` | [mod.md](mod.md) |
 | `options.js` | `option`, `Option`, `Options`, `OptionValues`, `OptionValue`, `OptionType`, `UserConfigField`, `userConfigOf` | [options.md](options.md) |
-| `shell.js` | `parseShell`, `ParsedShell`, `ShellCommand` | [hooks.md](hooks.md) |
-| `path.js` | `resolve`, `join`, `dirname`, `basename`, `extname`, `relative`, `isAbsolute`, `normalize` | [hooks.md](hooks.md) |
-| `testing.js` | `testMod`, `TestOptions`, `TestInput`, `TestedMod`, `Fakes`, `Shown`, `ShownToast`, `TestCall`, `FakeSession`, `MessageDraft` | [testing.md](testing.md) |
+| `shell.js` | `parseShell`, `ParsedShell`, `ShellCommand` | [mod.md](mod.md#shell-js-and-path-js-read-a-command-line) |
+| `path.js` | `resolve`, `join`, `dirname`, `basename`, `extname`, `relative`, `isAbsolute`, `normalize` | [mod.md](mod.md#shell-js-and-path-js-read-a-command-line) |
+| `testing.js` | `testMod`, `TestOptions`, `TestedMod`, `Raise`, `Fakes`, `Shown`, `ShownToast`, `TestCall`, `FakeSession`, `MessageDraft`, `RoutedEvent` | [testing.md](testing.md) |
 | `ui/define-pane.js` | `definePane`, `Pane` | [ui.md](ui.md) |
 | `ui/elements.js` | `Box`, `Text`, `Button`, `Link`, `Code`, `Markdown`, `Input`, `Select`, `Image`, `drawWith` | [ui.md](ui.md) |
 | `ui/components.js` | `Tabs`, `Tab`, `Split`, `Panel`, `Toggle`, `Tooltip`, `Help`, `Dialog`, `Pagination`, `Progress` | [ui.md](ui.md) |

package/docs/jobs.md

@@ -258,25 +258,22 @@ It needs `"prompt": true`. Without the grant it adds nothing and logs once which
 ```ts
 prompt<State>(options: {
   name: string
-  prompt: string | ((input: { userPrompt?: string; call?: Call & { folder: string } }, mod: Mod<State>) => string | undefined | Promise<string | undefined>)
-  when?: RegExp | ((userPrompt: string) => boolean)
+  prompt: string | ((input: { call?: Call & { folder: string } }, mod: Mod<State>) => string | undefined | Promise<string | undefined>)
   after?: Target | Target[]
 }): Job<void, State>
 ```
 
-`prompt` adds text that Claude reads. `when` and `after` pick when:
+`prompt` adds text that Claude reads. `after` picks when:
 
 | Options | When Claude reads the text |
 | --- | --- |
-| neither | At the start of each conversation, as a block of context named `name`. A block of that name already in the context wins. Then again whenever `mod.state` changes and the text differs from what Claude last got. |
-| `when` | With each prompt the person sends whose text matches the `RegExp`, or for which the function returns true. `input.userPrompt` is the prompt. |
-| `after` | After each call that matches a target and succeeds, with the call's result. `input.call` is the matched call, with its `folder` as `permissions` gives it. |
+| no `after` | At the start of each conversation, as a block of context named `name`. A block of that name already in the context wins. Then again whenever `mod.state` changes and the text differs from what Claude last got. |
+| `after` | After each call that matches a target and succeeds, with the call's result, under the heading `# <name>`. `input.call` is the matched call, with its `folder` as `permissions` gives it. |
 
-- With `when` or `after`, the text goes to Claude under the heading `# <name>`.
-- With neither, the text follows the mod's state. After a change to `mod.state`, cmod calls `prompt` again, and when the text differs from what Claude last got, it adds `# <name>` and the new text to the conversation as a `Note` Claude reads on its next call to the model, as `sessions.current.send` does ([session.md](session.md#send-to-a-conversation)). Several changes in a row give one note with the last text. A note waits until the conversation has started, so a state change before the first prompt changes only the first block.
+- Without `after`, the text follows the mod's state. After a change to `mod.state`, cmod calls `prompt` again, and when the text differs from what Claude last got, it adds `# <name>` and the new text to the conversation as a `Note` Claude reads on its next call to the model, as `sessions.current.send` does ([session.md](session.md#act-on-a-session)). Several changes in a row give one note with the last text. A note waits until the conversation has started, so a state change before the first prompt changes only the first block.
 - A `prompt` function that returns `undefined` or `''` adds nothing. One that throws adds nothing, and the error goes to Claude Code's debug log.
-- `name` cannot be blank. `when` and `after` together throw: add a second `prompt` for the other trigger. Two prompts with one name in a mod throw.
-- A `when` prompt runs on `prompt.submit` and adds its text to the event it passes down. In a test, `fire('prompt.submit', …)` returns that text in the answer's `context`. [testing.md](testing.md#test-a-prompt-with-when) shows how.
+- `name` cannot be blank. Two prompts with one name in a mod throw.
+- Text for the prompts that match a pattern is a `beforePrompt` note: `sessions.current.beforePrompt({ text: /\bmigrat/i }, () => ({ note: 'Every migration has a down step.' }))` ([session.md](session.md#change-what-happens)).
 
 ```ts
 import { defineMod, fs, sessions } from '../node_modules/@cmodjs/core/mod.js'
@@ -288,7 +285,6 @@ export const conventions = defineMod({
   setup(mod) {
     mod.use(prompt({ name: 'Commit messages', prompt: 'Write commit messages in the imperative mood.' }))
     mod.use(prompt({ name: 'Mode', prompt: (_input, mod) => `You are in ${mod.state.session.mode} mode.` }))
-    mod.use(prompt({ name: 'Migrations', when: /\bmigrat/i, prompt: 'Every migration has a down step.' }))
     mod.use(prompt({ name: 'Schema', after: { write: 'db/schema.sql' }, prompt: async () => `The schema is now:\n${await fs.read(`${sessions.current.project}/db/schema.sql`)}` }))
   },
 })
@@ -411,19 +407,19 @@ type JobContext<State> = {
 }
 ```
 
-- `mod` is the mod that uses the job. A job builds on `mod.on`, `mod.every`, `mod.ui`, and the imports from `mod.js`, as `setup` does ([mod.md](mod.md)).
+- `mod` is the mod that uses the job. A job builds on `mod.every`, `mod.ui`, and the imports from `mod.js`, such as the methods of `sessions`, as `setup` does ([mod.md](mod.md)).
 - `announce` adds a phrase to the list cmod logs when the mod first starts.
 - `reserveName` throws `taken` when the mod already reserved that `type` and `name`, so a job refuses a duplicate. `dispose()` on what it returns frees the name again.
 
 A job that hands back a feature the mod turns off returns a `Disposable`, as `slashCommand` and `tool` do:
 
 ```ts
-import { defineMod, sessions, Stop, ToolCall, type Disposable, type Job } from '../node_modules/@cmodjs/core/mod.js'
+import { defineMod, sessions, type Disposable, type Job } from '../node_modules/@cmodjs/core/mod.js'
 
 const editCounter = (tool: string): Job<Disposable & { readonly count: () => number }> => ({ mod, announce, reserveName }) => {
   const reserved = reserveName('editCounter', tool, `${mod.name}: edits of ${tool} are already counted.`)
   let edits = 0
-  const listening = sessions.current.on(ToolCall, { name: tool }, () => void (edits += 1))
+  const listening = sessions.current.onToolCall({ name: tool }, () => void (edits += 1))
   announce(`a count of ${tool} calls`)
   return {
     count: () => edits,
@@ -438,7 +434,7 @@ export const counter = defineMod({
   name: 'counter',
   setup(mod) {
     const writes = mod.use(editCounter('Write'))
-    mod.on(Stop, () => mod.ui.toast(`${writes.count()} files written so far`))
+    sessions.current.onTurnEnd(() => mod.ui.toast(`${writes.count()} files written so far`))
   },
 })
 ```
@@ -460,4 +456,4 @@ A job of your own matches calls against targets through the `permissions`, `chec
 
 ## Test a job
 
-`tested.type('/todo milk')` runs a slash command. `tested.callTool('open_ticket', { title: 'Crash' })` runs the permission rules, then the tool. `tested.fire('tool.check', …)` asks the permission rules about any call. `tested.fire('prompt.submit', …)` returns the context a `when` prompt added. [testing.md](testing.md) covers each.
+`tested.type('/todo milk')` runs a slash command. `tested.callTool('open_ticket', { title: 'Crash' })` runs the permission rules, then the tool. `tested.fire('tool.check', …)` asks the permission rules about any call. `tested.shown.notes` holds the notes a `prompt` sent after a state change. [testing.md](testing.md) covers each.

package/docs/mod.md

@@ -24,17 +24,17 @@ type ModDefinition<State, Name, Options> = {
 - `options` declares what a person sets, such as a token, and types `mod.options`. [options.md](options.md) explains them.
 - `api` holds the methods other mods call. [dependencies.md](dependencies.md) explains it.
 - `installer` lists the steps the person finishes in Claude Code after the permissions and the options, such as signing in. [install-steps.md](install-steps.md#steps-in-claude-code) explains them.
-- `setup` runs once when the mod starts. It adds hooks, panes, renders, and jobs. It may be `async`.
+- `setup` runs once when the mod starts. It adds listeners, panes, renders, and jobs. It may be `async`.
 
 ```tsx
-import { defineMod, UserPromptSubmit } from '../node_modules/@cmodjs/core/mod.js'
+import { defineMod, sessions } from '../node_modules/@cmodjs/core/mod.js'
 
 export const greeter = defineMod({
   name: 'greeter',
   state: { session: { isGreeted: false } },
 
   setup(mod) {
-    mod.on(UserPromptSubmit, () => {
+    sessions.current.onPrompt(() => {
       if (mod.state.session.isGreeted) return
       mod.state.session.isGreeted = true
       mod.ui.toast(`Hello from ${mod.name}`)
@@ -47,7 +47,7 @@ export const greeter = defineMod({
 
 Claude Code loads the mod at the start of each session. cmod then checks that the mod is set up. A mod the person has not accepted yet opens the installer pane, `Install <name>`, which lists its permissions and the changes it makes to the computer, and waits for Accept ([install-steps.md](install-steps.md)). An option with no default and no value is asked on the next page of the same pane. Then cmod loads the saved state, runs `setup`, marks as open each of the mod's panes that Claude Code already shows, and opens the pane again for each `installer` step the mod declares that is not done.
 
-The first time a mod starts on a machine, cmod shows the toast `Installed` under its name. Opened, it lists the mod's `plugin.json` `description`, what setup added, such as `Adds the my-mod pane, a render of AbovePrompt and a hook on UserPromptSubmit.`, what its install step set up, and each permission the person allowed, with a Settings button for `/mods <name>`. The first time a new version starts, the toast is `Updated to <version>`: the first three changes of that version's `CHANGELOG.md` section, the version the machine ran before, and a Changelog button ([Release notes](#release-notes)). A mod with no changelog lists what it adds now and no longer adds instead. A new session, `/reload-plugins`, and a version the machine already ran show no toast.
+The first time a mod starts on a machine, cmod shows the toast `Installed` under its name. Opened, it lists the mod's `plugin.json` `description`, what setup added, such as `Adds the my-mod pane, a render of AbovePrompt and /deploy.`, what its install step set up, and each permission the person allowed, with a Settings button for `/mods <name>`. The first time a new version starts, the toast is `Updated to <version>`: the first three changes of that version's `CHANGELOG.md` section, the version the machine ran before, and a Changelog button ([Release notes](#release-notes)). A mod with no changelog lists what it adds now and no longer adds instead. A new session, `/reload-plugins`, and a version the machine already ran show no toast.
 
 When a step fails, or the person declines the install, the mod stays stopped. Once another session sets it up, the mod starts at the next prompt of every session it stopped in, with no `/reload-plugins`.
 
@@ -59,7 +59,7 @@ When a step fails, the mod does not start. An error toast above the prompt names
 - `its setup function threw: <error>` when `setup` throws or rejects.
 - `its open panes did not load: <error>` when Claude Code cannot list the open panes.
 - `the details of its session did not load: <error>` when Claude Code cannot name the session's id, model, or start time.
-- `it decides permissions on <events>, so hooks/register.ts must call registerPermissionCheck(addHook) after registerMod` when `setup` adds a hook on `PermissionRequest`, or a job that decides permissions, and `register.ts` does not register the permission check.
+- `it decides permissions on <events>, so hooks/register.ts must call registerPermissionCheck(addHook) after registerMod` when `setup` adds a `beforePermissionCheck` or `beforePermissionRequest` listener, or a job that decides permissions, and `register.ts` does not register the permission check.
 
 Each of these has `Retry`, which starts the mod again, and `Report` and `Hide` ([ui.md](ui.md#toasts)).
 
@@ -123,9 +123,9 @@ registerPermissionCheck(addHook: On): void
 registeredMod(): ModDefinition | undefined
 ```
 
-`registerMod` registers one handler per Claude Code event the mod can use, and routes each event to the mod. `options` is what Claude Code passes `register`: the values of the mod's options ([options.md](options.md)). `registeredMod` returns the definition `registerMod` got, for `cmod check` to read the mod's options without starting it. A `PreToolUse` hook runs on `tool.call`, which `registerMod` registers ([hooks.md](hooks.md)). It leaves out Claude Code's permission check, the events `tool.check` and `PermissionRequest`.
+`registerMod` registers one handler per Claude Code event the mod can use, and routes each event to the mod. `options` is what Claude Code passes `register`: the values of the mod's options ([options.md](options.md)). `registeredMod` returns the definition `registerMod` got, for `cmod check` to read the mod's options without starting it. `registerMod` leaves out Claude Code's permission check, the events `tool.check` and `PermissionRequest`.
 
-A mod that decides permissions calls `registerPermissionCheck` after `registerMod`. That covers a `PreToolUse` hook that answers `permissionDecision: 'allow'` or `'ask'`, a hook on `PermissionRequest`, the `permissions` job, and a job of your own on `tool.check`:
+A mod that decides permissions calls `registerPermissionCheck` after `registerMod`. That covers `sessions.current.beforePermissionCheck`, `sessions.current.beforePermissionRequest`, and the `permissions` job:
 
 ```ts
 export function register(addHook: On, options: PluginOptions): void {
@@ -136,11 +136,11 @@ export function register(addHook: On, options: PluginOptions): void {
 
 Such a mod also declares `"approve": true` in its `package.json` `"cmod".permissions`, so the person grants it at install. `cmod check` adds the line when it is missing, and without the grant cmod drops every allow or ask the mod answers, so Claude Code decides those calls itself ([permissions.md](permissions.md)). A mod without `registerPermissionCheck` has no part in Claude Code's permission check, so the plugin directory reads it as one that never answers a permission. A mod that decides permissions without it does not start, and its error toast names the line to add.
 
-Keep `register.ts` this small. Claude Code checks a hooks module before it loads it, and refuses some shapes, such as a `$` passed to a function in another module. The plugin directory reads the bundled module too, and flags it when the name of `register`'s first parameter is declared anywhere else in the bundle. `@cmodjs/core` declares `on` as the method `mod.on`, so name that parameter `addHook`. Declare `register` as a function, as here: `cmod publish` bundles the hooks module into one file, and Claude Code refuses a bundled `register` that is not one.
+Keep `register.ts` this small. Claude Code checks a hooks module before it loads it, and refuses some shapes, such as a `$` passed to a function in another module. The plugin directory reads the bundled module too, and flags it when the name of `register`'s first parameter is declared anywhere else in the bundle. `@cmodjs/core` declares `on` as a key, so name that parameter `addHook`. Declare `register` as a function, as here: `cmod publish` bundles the hooks module into one file, and Claude Code refuses a bundled `register` that is not one.
 
 ## mod: this mod
 
-`setup` gets `mod`, of type `Mod<State, Options>`. Every hook, render, and job gets the same `mod`. `mod` holds what belongs to this mod: its name, its folders, its values, its grants, and what it adds to Claude Code.
+`setup` gets `mod`, of type `Mod<State, Options>`. Every listener, render, and job gets the same `mod`. `mod` holds what belongs to this mod: its name, its folders, its values, its grants, and what it adds to Claude Code.
 
 ```ts
 type Mod<State, Options> = {
@@ -150,8 +150,6 @@ type Mod<State, Options> = {
   readonly state: Readonly<State>
   readonly options: OptionValues<Options>
   readonly permissions: Permissions
-  on(hook, listener, options?): Disposable
-  on(hook, matcher, listener, options?): Disposable
   every(ms, listener): Disposable
   after(ms, listener): Disposable
   use(job): Handle
@@ -170,7 +168,6 @@ type Mod<State, Options> = {
 | `state` | The mod's values. Assign to a key to change it. | [state.md](state.md) |
 | `options` | The values a person set for the mod's options, typed from their declaration. Read-only. | [options.md](options.md) |
 | `permissions` | Says whether the person granted each permission. | [permissions.md](permissions.md) |
-| `on` | Listens to a Claude Code hook, such as `PreToolUse`. | [hooks.md](hooks.md) |
 | `every` and `after` | Run a listener on a timer while the session runs. | below |
 | `use` | Adds a job, such as a slash command or a tool, and returns its handle. | [jobs.md](jobs.md) |
 | `log` | Writes a line, under the mod's name, to Claude Code's debug log. | below |
@@ -181,13 +178,15 @@ type Mod<State, Options> = {
 Everything that is Claude Code's or the computer's rather than the mod's is an import from `mod.js`, so a function in any file calls it without being handed `mod`:
 
 ```ts
-import { fs, http, metadata, model, process, sessions, settings } from '../node_modules/@cmodjs/core/mod.js'
+import { commands, config, fs, http, metadata, model, process, sessions, settings } from '../node_modules/@cmodjs/core/mod.js'
 ```
 
 | Import | What it is | Docs |
 | --- | --- | --- |
-| `sessions` | The conversations: the one the mod runs in, every other one, and search across them. | [session.md](session.md) |
+| `sessions` | The conversations: the one the mod runs in, every other one, and search across them. Everything that happens in a session, from a prompt to a permission dialog, is a method on it. | [session.md](session.md) |
+| `commands` | Every slash command: lists them, runs one, and hears or changes each run. | below |
 | `settings` | Claude Code's `settings.json` files: the person's, the project's, and the project's local one. | below |
+| `config` | The rows of Claude Code's `/config` menu. | below |
 | `model` | Asks a model one question, outside the conversation. | below |
 | `fs` | Reads, writes, lists, and finds files. | below |
 | `metadata` | Reads and changes the mod's keys in one file's metadata. | below |
@@ -198,14 +197,14 @@ Each import works once the mod has started, in `setup` and anything it adds. Eac
 
 ### Disposable
 
-`mod.on`, `mod.every`, `mod.after`, `mod.ui.render`, and a session's `on` each return a `Disposable`. `dispose()` takes back what the call added, so a mod turns a feature off and on while it runs. Everything a mod adds is taken back when the mod stops, so a mod that keeps a feature for the whole session ignores the `Disposable`.
+`mod.every`, `mod.after`, `mod.ui.render`, and every `on` and `before` method each return a `Disposable`. `dispose()` takes back what the call added, so a mod turns a feature off and on while it runs. Everything a mod adds is taken back when the mod stops, so a mod that keeps a feature for the whole session ignores the `Disposable`.
 
 ```ts
 type Disposable = { dispose(): void }
 ```
 
 ```ts
-const watching = mod.on(PostToolUse, { tool_name: 'Edit' }, () => void (mod.state.session.edits += 1))
+const watching = sessions.current.onToolCall({ name: 'Edit' }, () => void (mod.state.session.edits += 1))
 mod.use(slashCommand({ name: 'stop-counting', description: 'Stop counting edits', reply: () => watching.dispose() }))
 ```
 
@@ -230,11 +229,13 @@ mod.every(60_000, async () => {
 ### mod.log
 
 ```ts
-mod.log(text: string): void
+mod.log(text: string, options?: { to?: 'debug' | 'transcript' }): void
 ```
 
 `log` writes `<mod>: <text>` to Claude Code's debug log, which `claude --debug` shows. It is for what helps the mod's author find a fault and the person never needs to see. Tell the person with a toast instead ([ui.md](ui.md#toasts)).
 
+With `{ to: 'transcript' }`, the line shows dim in the conversation, where the person reads it and Claude never does, such as "Formatted 3 files" after a turn.
+
 ### mod.settings
 
 ```ts
@@ -249,26 +250,43 @@ mod.settings.page(definePane({ id: 'runs', title: 'Runs', render: (current) => <
 mod.use(slashCommand({ name: 'ci-runs', description: 'Show the CI runs', reply: () => void mod.settings.open('runs') }))
 ```
 
+## commands: slash commands
+
+```ts
+commands.list(): Promise<CommandInfo[]>
+commands.run(name: string, args?: string): Promise<{ text?: string; note?: string }>
+commands.beforeRun(matcher?, listener: (run: CommandRun) => Answer | void, options?): Disposable
+commands.onRun(matcher?, listener: (run: CommandRun) => unknown, options?): Disposable
+
+type CommandInfo = { name: string; description: string; source: 'builtin' | 'plugin' | 'user' | 'mcp'; plugin?: string }
+type CommandRun = { name: string; args: string; source?: 'builtin' | 'plugin' | 'user' | 'mcp'; from: Sender; session: string; cwd: string }
+```
+
+- `list` answers every command the person can type: Claude Code's own, each plugin's, the person's and the project's, and each MCP server's prompts.
+- `run` runs a command, a built-in such as `compact` included, the way the person typing `/<name> <args>` would. It resolves the text the command printed and the note it left Claude. It needs the `prompt` permission.
+- `beforeRun` runs just before any command runs, typed or run by a mod, and its answer changes it: `{ args?, note? }` runs it with new arguments and adds a note Claude reads after its output, `{ result, exitCode? }` answers in its place without running it, and `{ refuse }` runs nothing and shows the reason as its output. An answer needs the `prompt` permission. `exitCode` is what a `claude -p "/lint"` run exits with.
+- `onRun` hears each command once it ran.
+- A command a project or a plugin writes as a Markdown prompt, such as `/review`, also reaches `beforePrompt` after `beforeRun`, as its text goes to Claude ([session.md](session.md#change-what-happens)). A built-in such as `/model` reaches only these two.
+
+```ts
+commands.beforeRun({ name: 'compact' }, () => (mod.state.session.isDeploying ? { refuse: 'Compact once the deploy finishes.' } : undefined))
+commands.onRun({ name: 'model' }, (run) => mod.log(`model set to ${run.args}`))
+```
+
+A command of the mod's own is a `slashCommand` job ([jobs.md](jobs.md#slashcommand)).
+
 ## settings: Claude Code's settings files
 
 ```ts
-settings.list(): Promise<Setting[]>
 settings.user.read(): Promise<Settings>
 settings.user.update(values: Readonly<Record<string, unknown>>): Promise<void>
 settings.project.read(): Promise<Settings>
 settings.project.update(values: Readonly<Record<string, unknown>>): Promise<void>
 settings.local.read(): Promise<Settings>
 settings.local.update(values: Readonly<Record<string, unknown>>): Promise<void>
+settings.beforeFileChange(matcher?, listener: (change: SettingsFileChange) => { refuse: string } | void): Disposable
 ```
 
-`settings.list` answers every row of Claude Code's `/config` menu, its own and each plugin's, with the value each holds now:
-
-```ts
-type Setting = { key: string; label: string; description?: string; kind: 'boolean' | 'choice' | 'text' | 'number'; value: boolean | string | number | readonly string[]; options?: readonly string[]; plugin?: string; isLocked: boolean }
-```
-
-`plugin` names the plugin a row belongs to, and Claude Code's own rows have none. `isLocked` is true when a managed setting or the organization's policy holds the value.
-
 `settings.user` is the person's own `~/.claude/settings.json`, `settings.project` is the project's `.claude/settings.json`, and `settings.local` is the project's `.claude/settings.local.json`.
 
 - `read` answers the values in that one file, such as `(await settings.user.read())['enabledPlugins']`.
@@ -285,6 +303,24 @@ A mod that points Claude Code at its local proxy sets one variable in `env`, and
 await settings.user.update({ env: { ANTHROPIC_BASE_URL: 'http://127.0.0.1:8317' } })
 ```
 
+`beforeFileChange` hears each settings file Claude Code sees edited while the session runs, and a Skill file edited, and `{ refuse }` keeps Claude Code on the values it had. `source` is `'user'`, `'project'`, `'local'`, `'policy'`, or `'skills'`, and `path` is the file, when Claude Code names it:
+
+```ts
+type SettingsFileChange = { source: 'user' | 'project' | 'local' | 'policy' | 'skills'; path?: string }
+
+settings.beforeFileChange({ source: /^(user|project)$/ }, () => ({ refuse: 'Settings are locked while the deploy runs.' }))
+```
+
+## config: the /config menu
+
+```ts
+config.list(): Promise<ConfigRow[]>
+
+type ConfigRow = { key: string; label: string; description?: string; kind: 'boolean' | 'choice' | 'text' | 'number'; value: boolean | string | number | readonly string[]; options?: readonly string[]; plugin?: string; isLocked: boolean }
+```
+
+`config.list` answers every row of Claude Code's `/config` menu, its own and each plugin's, with the value each holds now. `plugin` names the plugin a row belongs to, and Claude Code's own rows have none. `isLocked` is true when a managed setting or the organization's policy holds the value.
+
 ## model: one question outside the conversation
 
 ```ts
@@ -295,12 +331,12 @@ model.complete(request: ModelCompleteRequest, options?: { signal?: AbortSignal }
 
 The call never rejects for what the model did. The result's `isAnswered` is true with the reply's `text`, or false with a `reason`, such as `'api-error'` or `'aborted'`.
 
-Put a limit on the listener too, with `mod.on`'s `timeoutMs` ([hooks.md](hooks.md#mod-on)), which also covers the work around the call:
+Give the listener that waits on it time too, with its `timeoutMs` ([session.md](session.md#give-a-listener-time)), which also covers the work around the call:
 
 ```ts
-mod.on(Stop, async ({ last_assistant_message }) => {
-  const verdict = await model.complete({ model: 'haiku', prompt: `Does this reply ask the person to do work Claude could do itself? Answer yes or no.\n\n${last_assistant_message}`, timeoutMs: 8000 })
-  if (verdict.isAnswered && verdict.text.trim().toLowerCase().startsWith('yes')) return { decision: 'block', reason: 'Do that work yourself, then reply.' }
+sessions.current.beforeTurnEnd(async (turn, { signal }) => {
+  const verdict = await model.complete({ model: 'haiku', prompt: `Does this reply ask the person to do work Claude could do itself? Answer yes or no.\n\n${turn.answer ?? ''}`, timeoutMs: 8000 }, { signal })
+  if (verdict.isAnswered && verdict.text.trim().toLowerCase().startsWith('yes')) return { refuse: 'Do that work yourself, then reply.' }
 }, { timeoutMs: 9000 })
 ```
 
@@ -325,15 +361,15 @@ Both run a program without a shell. `argv[0]` is the program and the rest are it
 A job's deadline bounds the whole callback that makes these calls ([jobs.md](jobs.md#deadlines)).
 
 ```ts
-import { defineMod, process, SessionStart, sessions } from '../node_modules/@cmodjs/core/mod.js'
+import { defineMod, process, sessions } from '../node_modules/@cmodjs/core/mod.js'
 
 export const branchNote = defineMod({
   name: 'branch-note',
-  setup(mod) {
-    mod.on(SessionStart, async () => {
+  setup() {
+    sessions.current.beforeSessionStart(async () => {
       const { exitCode, stdout } = await process.run(['git', 'branch', '--show-current'], { cwd: sessions.current.project })
       if (exitCode !== 0) return
-      return { hookSpecificOutput: { hookEventName: SessionStart, additionalContext: `The git branch is ${stdout.trim()}.` } }
+      return { note: `The git branch is ${stdout.trim()}.` }
     })
   },
 })
@@ -348,9 +384,11 @@ fs.list(path?: string): Promise<Entry[]>
 fs.exists(path: string): Promise<boolean>
 fs.stat(path: string, options?: { resolve: boolean }): Promise<Stat>
 fs.find(glob: string | readonly string[]): Promise<readonly FileMatch[]>
+fs.onChange({ path: string | readonly string[] }, listener: (change: FileChange) => unknown): Disposable
 ```
 
 ```ts
+type FileChange = { path: string; type: 'change' | 'add' | 'unlink' }
 type Entry = { name: string; isFile: boolean; isFolder: boolean; isLink: boolean; size: number; modifiedAt: number }
 type Stat = { isFile: boolean; isFolder: boolean; isLink: boolean; size: number; modifiedAt: number; realPath?: string }
 ```
@@ -361,6 +399,11 @@ type Stat = { isFile: boolean; isFolder: boolean; isLink: boolean; size: number;
 - `list` lists a folder, the working folder without a path. A symbolic link is neither a file nor a folder, with `isLink` true. `modifiedAt` is the time it last changed, in milliseconds since 1970.
 - `exists` answers whether the path leads to a file or a folder.
 - `stat` answers what the path leads to, following a symbolic link, and rejects a missing path. With `{ resolve: true }` it also answers `realPath`: the absolute path, every symbolic link followed.
+- `onChange` hears each file Claude Code sees change, appear, or go at a path or under a folder. Claude Code reads the paths it watches when a session starts, so `onChange` works only in `setup`, and a call later throws. A relative path is under the project. `path` takes strings only, unlike other matchers, because the paths go to Claude Code as they are:
+
+```ts
+fs.onChange({ path: `${mod.dataFolder}/history` }, ({ path, type }) => mod.ui.toast(`${path}: ${type}`))
+```
 
 A file the mod ships is under `mod.folder`. A file the mod keeps for itself belongs in `mod.dataFolder`.
 
@@ -443,12 +486,12 @@ messageOf(error: unknown): string
 Returns an `Error`'s `message`, or the value as text. Use it in a `catch`, where the error is `unknown`.
 
 ```ts
-import { defineMod, fs, messageOf, Stop } from '../node_modules/@cmodjs/core/mod.js'
+import { defineMod, fs, messageOf, sessions } from '../node_modules/@cmodjs/core/mod.js'
 
 export const notes = defineMod({
   name: 'notes',
   setup(mod) {
-    mod.on(Stop, async () => {
+    sessions.current.onTurnEnd(async () => {
       await fs.write(`${mod.dataFolder}/last-stop.txt`, new Date().toISOString()).catch((error: unknown) => {
         mod.ui.toast.error('Could not save the last stop', { description: messageOf(error) })
       })
@@ -456,3 +499,35 @@ export const notes = defineMod({
   },
 })
 ```
+
+## shell.js and path.js: read a command line
+
+`shell.js` reads a Bash line the way cmod does for a tool call's `files` ([session.md](session.md#files-what-a-tool-call-reads-and-changes)), and `path.js` is `node:path`'s POSIX functions, which a mod cannot import from Node:
+
+```ts
+parseShell(line: string): ParsedShell
+
+type ParsedShell = { commands: ShellCommand[]; reads: string[]; writes: string[]; fetches: string[]; isFullyParsed: boolean }
+type ShellCommand = { argv: [string, ...string[]]; folder: string; input?: number }
+
+resolve(...paths): string   join(...paths): string   dirname(path): string   basename(path, suffix?): string
+extname(path): string        relative(from, to): string   isAbsolute(path): boolean   normalize(path): string
+```
+
+- `commands` lists every command the line runs, those inside `sudo`, `bash -c`, `$(…)`, and heredocs included, with flags such as `-rf` split into `-r` and `-f`.
+- `folder` is where a command runs, as the line wrote it: `''` for the folder the line starts in, `src` after `cd src`, `/tmp` after `cd /tmp`, and `~/p` after `cd ~/p`. `resolve(call.cwd, folder)` makes it absolute.
+- `input` is the index in `commands` of the command whose output this one reads through `|` or `|&`, and is absent for a command that reads no pipe. `trace read a.ts | head -5` gives `head` the `input` `0`. A wrapper such as `timeout 5 trace read a.ts` writes the output of the command it runs, so `input` names `trace`, not `timeout`.
+- `reads`, `writes`, and `fetches` are the paths and URLs the commands name, as written. A path that holds a shell expansion makes `isFullyParsed` false, and so does a line the parser cannot read.
+- `resolve` has no working folder to start from in Claude Code, so pass an absolute folder first, such as `resolve(sessions.current.cwd, path)`.
+
+```ts
+import { basename, resolve } from '../node_modules/@cmodjs/core/path.js'
+import { parseShell } from '../node_modules/@cmodjs/core/shell.js'
+import { sessions } from '../node_modules/@cmodjs/core/mod.js'
+
+sessions.current.onToolCall({ name: 'Bash' }, (call) => {
+  for (const { argv, folder } of parseShell(String(call.input['command'])).commands) {
+    if (basename(argv[0]) === 'trace') mod.ui.toast(`trace runs in ${resolve(call.cwd, folder)}`)
+  }
+})
+```