Editing pass over the docs and code comments before the v0.2.0 release

.claude/skills/verify/SKILL.md

@@ -43,8 +43,9 @@ dispatch = "echo usage limit reached # {prompt_file}" # cap phrase -> capped
 ```
 
 The dispatch template must contain `{prompt_file}` or config validation
-rejects it. Capped detection matches "usage limit", "rate limit", or
-"quota exceeded" in the log tail (`crates/voro/src/reconcile.rs`).
+rejects it. Capped detection matches the signatures in
+`crates/voro-core/src/cap.rs` (`CAP_SIGNATURES` — "usage limit",
+"session limit", "weekly limit", …) against the log tail.
 
 ## Driving the TUI
 
@@ -59,8 +60,9 @@ broken for non-interactive use — call `/bin/tmux` directly:
 /bin/tmux -L <socket> kill-server
 ```
 
-`r` refreshes in place after CLI mutations from outside; `PAGER` is honoured
-by the `l` log key, so setting it to `less` keeps the pager capturable.
+`ctrl-r` refreshes in place after CLI mutations from outside (`r` is the
+refine key); `PAGER` is honoured by the `l` log key, so setting it to `less`
+keeps the pager capturable.
 
 A scratch store opens empty — the fixture seeds only the dev store — so the
 first capture of an unseeded run shows a board with no projects and no tasks.

README.md

@@ -5,7 +5,7 @@
 [![docs.rs](https://img.shields.io/docsrs/voro-core)](https://docs.rs/voro-core)
 [![License](https://img.shields.io/badge/license-MIT%20OR%20Apache--2.0-blue.svg)](LICENSE-MIT)
 
-An attention based session manager for AI-assisted development across many
+An attention-based session manager for AI-assisted development across many
 projects. Voro tracks tasks per project, weights each project by how much it
 matters *today*, and answers one question: **where should your attention go right
 now?**

crates/voro-core/src/agent.rs

@@ -105,14 +105,10 @@ pub const VIEWER_BASE_PLACEHOLDER: &str = "{base}";
 /// deliberate: a verb is an opaque per-agent contract, which is exactly what
 /// lets an agent define a subset of them and degrade per-verb. `codex` defines
 /// no `message` and the TUI's quick-message key says so on the status line.
-/// It resumes the session in place rather than forking it
-/// ([`NEW_SESSION_PLACEHOLDER`], which the verb no longer carries): a
-/// `claude --bg` session keeps its supervisor process after finishing its turn,
-/// and that supervisor refuses a headless `--resume` for as long as it lives —
-/// so Voro releases it at rest instead, through `stop`, and the send then lands
-/// on the session's own reference (DESIGN.md §8). A fork would land too, but it
-/// moves the conversation out from under the name Voro composed for it, and that
-/// name is how the operator addresses the session everywhere outside Voro.
+/// A `claude --bg` session keeps its supervisor process after finishing its
+/// turn, and that supervisor refuses a headless `--resume` for as long as it
+/// lives — Voro releases it at rest through `stop`, so the send lands on the
+/// session's own reference (DESIGN.md §8).
 ///
 /// It carries `--permission-mode` for the same reason `dispatch` does: the mode
 /// belongs to a launch rather than to a verb (DESIGN.md §8). The flag is per
@@ -1938,7 +1934,7 @@ mod tests {
         }
     }
 
-    // --- plan verb (task #112) ---
+    // --- plan verb ---
 
     #[test]
     fn plan_parses_resolves_and_is_optional() {
@@ -2386,9 +2382,9 @@ mod tests {
         assert_eq!(rendered.command, "say '{prompt_file}' '/run/m.md'");
     }
 
-    /// A stale `continue` line — from a pre-pivot config or the old codex
-    /// built-in — is now an unknown field, so the config is refused rather than
-    /// silently honouring a verb Voro no longer runs (DESIGN.md §6/§8).
+    /// A `continue` line is an unknown field, so a config carrying one is
+    /// refused rather than loading with a verb Voro never runs (DESIGN.md
+    /// §6/§8).
     #[test]
     fn a_continue_verb_is_now_an_unknown_field() {
         let text = r#"
@@ -2511,8 +2507,8 @@ mod tests {
     /// Nothing installed, nothing configured: the failure asks for the one
     /// thing the operator can act on — register the viewer they already use —
     /// and only then says what was probed. It never tells them to install an
-    /// editor, and never calls the config file invalid: it may not even exist
-    /// (#405).
+    /// editor, and never calls the config file invalid: it may not even
+    /// exist.
     #[test]
     fn viewer_resolution_errors_with_guidance_when_nothing_resolves() {
         let message = config()
@@ -2812,7 +2808,7 @@ mod tests {
         assert!(agents["codex"].resume().is_some());
     }
 
-    // --- launch identity and rendered commands (task #326) ---
+    // --- launch identity and rendered commands ---
 
     /// A dispatch of task 7 with a fixed prompt file, so a rendered command is
     /// a stable string to assert on.
@@ -3228,8 +3224,8 @@ mod tests {
         );
     }
 
-    /// What a headless launch records on its session row (DESIGN.md §8, task
-    /// #387): an agent with a `sessions` verb may hand the work to a supervisor,
+    /// What a headless launch records on its session row (DESIGN.md §8): an
+    /// agent with a `sessions` verb may hand the work to a supervisor,
     /// so its listing is the authority; one without has only the pid Voro
     /// spawned.
     #[test]

crates/voro-core/src/config_edit.rs

@@ -383,7 +383,7 @@ cmd = \"git -C {path} difftool -d {base}...{branch}\"  # inline note
 
     /// A built-in viewer lives in the binary, so the write helpers cannot touch
     /// it — but they can be told to override it, and it is a legitimate default
-    /// with no table of its own (#405).
+    /// with no table of its own.
     #[test]
     fn built_in_viewers_are_unwritable_but_overridable_and_defaultable() {
         let dir = scratch("builtin");
@@ -446,7 +446,7 @@ cmd = \"git -C {path} difftool -d {base}...{branch}\"  # inline note
 
     /// Naming the editor is enough: the command a new operator has no way to
     /// guess is assumed, and naming a built-in starts from what it replaces
-    /// rather than from a worse guess at the same thing (#405).
+    /// rather than from a worse guess at the same thing.
     #[test]
     fn a_viewer_added_with_no_command_gets_the_obvious_one() {
         let dir = scratch("assumed");

crates/voro-core/src/store.rs

@@ -1600,7 +1600,7 @@ impl Store {
         self.session(id)
     }
 
-    /// Record the agent's own reference for a session (task #75), captured
+    /// Record the agent's own reference for a session, captured
     /// after launch — the row necessarily exists before the reference does,
     /// so this is an update rather than a `create_session` parameter.
     pub fn set_session_ref(&mut self, id: i64, session_ref: &str) -> Result<Session> {
@@ -1716,7 +1716,7 @@ impl Store {
     /// `refining`, or `waiting` task, joined with its open session if it has
     /// one. The strip filters on task *state*, so `review`/`needs-input` tasks
     /// (session still open) do not appear, while a refine in flight and a
-    /// handed-off task both do — an open session no longer implies executing
+    /// handed-off task both do — an open session does not imply executing
     /// the task (§8), and what the strip shows is work in flight that someone
     /// else owns. A hand-started task with no session shows with `session_id`/
     /// `agent` `NULL`. The one-open-session invariant (§8) bounds the join to one
@@ -2787,7 +2787,7 @@ mod tests {
         );
     }
 
-    // --- the deep flag (task #241) ---
+    // --- the deep flag ---
 
     fn deep_new(project_id: i64, human: bool, deep: bool) -> NewTask {
         NewTask {
@@ -4329,7 +4329,7 @@ mod tests {
     }
 
     /// Each launch records which source reconciliation must read it by
-    /// (DESIGN.md §8, task #387), and it survives the round trip: a headless
+    /// (DESIGN.md §8), and it survives the round trip: a headless
     /// launch under a supervisor is listing-authoritative, an interactive round
     /// is not, and neither is inferred from anything else on the row.
     #[test]