There still seems to be a TUI-sync issue on upstream/main where
the assistant cuts off mid-sentence on screen even though the
DB has the full response.  Two recent PRs probably get most of
the way to fixing it: #2840 reordered the agent's
TypeAgentFinished notification past the activeRequests cleanup,
and #2836 made channelBufferSize actually honored.  In
power-user cases the default (64) is still too small, though.

A streaming turn calls messages.Update on every OnTextDelta /
OnReasoningDelta callback, and the TUI's glamour re-render
periodically stalls the Update loop for tens to hundreds of
milliseconds at a time.  64 slots fill before drain catches up
and the non-blocking publish in broker.go silently drops the
rest.

This diff adopts a pretty aggressive fix -- 64 -> 4096 covers a
long turn at typical SSE rates even under pathological View()
stalls.  Cost is a few MB of ring-buffer allocation across all
brokers, comfortably inside any modern dev machine's budget;
users on tighter memory targets can still tune via
NewBrokerWithOptions.

Drive-by cleanup: maxEvents (struct field + NewBrokerWithOptions
param) is set to 1000 by default but never read anywhere in
tree -- it's bufferSize, not maxEvents, that actually bounds the
queue.  Drop maxEvents on the theory that misleading dead code
is worse than no code.

💘 Generated with Crush

Assisted-by: Claude Opus 4.7 via Crush <crush@charm.land>

Move notification publish inside requestMu lock to prevent inconsistent
UI state, and replace the sessionPermissions slice with csync.Map
for constant-time persistent permission lookups.

💘 Generated with Crush

Assisted-by: DeepSeek V4 Pro via Crush <crush@charm.land>

Co-authored-by: Tai Groot <tai@taigrr.com>

chore: harden directory bounds

Co-Authored-By: Charm Crush <crush@charm.land>

The upward search for crush.json had no boundary and would walk
all the way to the filesystem root, which meant a stray crush.json
placed high in the tree could be picked up by every project beneath
it. This applies the same project boundary used for .crush discovery.

Co-Authored-By: Charm Crush <crush@charm.land>

When Crush starts up it looks for an existing .crush directory by
walking from the working directory toward the filesystem root. In
layouts with several worktrees of the same project under a common
parent, that walk could end up putting crush.db files in the project
root in worktrees.

This stops the walk at the git working tree root, when one can be
detected, and otherwise at the working directory itself. Each project
should now get its own .crush as expected.

Co-Authored-By: Charm Crush <crush@charm.land>

Add variants of the upward filesystem search that stop at a
caller-supplied boundary directory instead of walking all the way up
to the filesystem root or to $HOME. Callers that want to avoid adopting
matches from outside their project can now express that boundary
explicitly. Existing callers and behavior are unchanged.

Co-Authored-By: Charm Crush <crush@charm.land>

Co-authored-by: Charm Crush <crush@charm.land>

We were not using the correct format. According to the SPEC it should be:

    Assisted-by: AGENT_NAME:MODEL_VERSION

So now it'll look like this:

    Assisted-by: Crush:kimi-k2.6

See:

https://docs.kernel.org/process/coding-assistants.html#attribution

I keep hitting a bug where a turn finishes -- desktop notification
fires, agent is clearly done -- but the terminal stays stuck on
the spinner with input blocked until something unrelated wakes
the TUI up.

The TUI polls IsSessionBusy() inside its TypeAgentFinished
handler and only re-evaluates on incoming tea.Msgs.  Run() right
now publishes the notification while activeRequests still holds
the session's cancel func, so the poll on receipt sees true and
the busy gates stay stuck.  Reorder so cleanup runs first.

💘 Generated with Crush

Assisted-by: Claude Opus 4.7 via Crush <crush@charm.land>

* Ref https://github.com/charmbracelet/x/pull/853 by @sven2718

Mirrors the view.go sanitizer pattern (filepath.Abs + filepath.Rel +
".." check) so CodeQL no longer flags os.Stat/MkdirAll/OpenFile in the
touch tool as uncontrolled path expressions.

feat(hooks): use embedded shell by default

config: lenient shell expansion default, uniform coverage across MCP, LSP, and providers

fix(config): restore full shell expansion in MCP config values

The assistant message type's toggle-expand method returned no value,
so it silently failed to satisfy the expandable interface and the
keyboard expand path skipped assistant thinking blocks. Expansion
only worked via the mouse, which called the concrete method directly.

Fixing the signature exposed a second bug: the mouse path then
toggled twice per click (once inside the item, once via the generic
interface path), netting no change. The item now just reports
whether the click hit the thinking box; the generic path does the
toggle, gated on that flag so clicks elsewhere do nothing.

Adds a runtime regression test plus a compile-time interface
assertion so this can't silently come back.

Co-authored-by: Crush <crush@charm.land>

When a tool call is made with invalid arguments, Crush will now report
the error back so the model will continue working and can do another
attempt.

Fix on Fantasy done by @mkaaad.

* Ref https://github.com/charmbracelet/crush/issues/2776
* Ref https://github.com/charmbracelet/fantasy/pull/223