LspFileSystem.write received the already-resolved absolute path, so
isInternalUrlPath could never see the original scheme ('/Users/theo/.omp/agent/sessions/-Projects-oh-my-pi/2026-06-10T09-11-41-506Z_019eb0cd-3ec2-7000-92aa-1b82aa4d78f0/local/,' vault://)
and the bridge guard only caught the active plan file via the secondary
absolutePath comparison.
Fix: store the original user-provided path in LspFileSystem as
requestedPath and pass it as the first arg to routeWriteThroughBridge.
This matches replace.ts and hashline/filesystem.ts which already pass
the unresolved path correctly.
Also remove the ReturnType<typeof spyOn> annotation from makeBridge()
per the AGENTS.md "NEVER use ReturnType<>" rule.
Addresses all three review concerns from @roboomp and @chatgpt-codex-connector:
1. **Contract regression fixed**: guard
(extracted from ) is now called by all four write sites.
Internal-URL paths (e.g. ) and the active plan file
in plan mode are never routed to the editor bridge.
2. **Bridge call hygiene**: bridge calls are now wrapped in ,
is called, and
is bumped — matching the contract already had.
3. **Tests**: adds 6 parity tests
(2 per write site) mirroring :
- routes plain workspace writes through the bridge, skips writethrough
- writes local plan artifacts to disk instead of the ACP bridge
**Shared module**: exports
and (guard +
bridge call + ToolError wrap + invalidateFsScanAfterWrite +
bumpFileMutationVersion). All four write sites call it; the duplicated
6-line pattern is gone. is simplified to use the same helper
instead of its private method.
Anthropic rejects tool_result blocks when is_error is true and content is
empty after trimming whitespace. Fill a placeholder at encode time and in
coerceToolResult so wedged sessions recover on the next request.
When Zed (or another ACP client) advertises the fs.writeTextFile
capability, the write tool already routes through it so the editor's
open buffer is updated immediately. The edit, patch, and replace modes
were missing this path — they always wrote directly to disk via the LSP
writethrough, leaving open buffers stale and requiring a workspace
reload before Zed's TypeScript diagnostics panel would reflect the
changes.
All three modes now check for an ACP bridge with writeTextFile support
and route through it when available, falling back to the existing LSP
writethrough path otherwise.
When a pattern like "qwen/qwen3-coder:exacto" contains "/" but no
recognised provider prefix, findExactModelReferenceMatch fails (it
treats "qwen" as the provider). Fall through to scan available models
by id so the pattern still matches the openrouter-hosted model.
Adds a regression test covering this case.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Extension commands (e.g. /sonnet) and TypeScript custom commands that
consume the input without calling the LLM return early from
session.prompt() with no agent turn. In ACP mode this left the pending
prompt promise unresolved, hanging the client forever.
Change session.prompt() to return Promise<boolean>: true when the LLM
was invoked, false when the command was fully handled locally.
#runPromptOrCommand calls #finishPrompt immediately on a false return so
the ACP turn completes.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
ACP_BUILTIN_SLASH_COMMANDS only carries primary names; the reserved set
passed to getRegisteredCommands was therefore missing aliases like
"models" (/model) and "force:" (/force). An extension registering one
of these aliases would appear in the palette but the builtin would win
at dispatch time (lookupBuiltinSlashCommand searches aliases too).
Export ACP_BUILTIN_RESERVED_NAMES from acp-builtins — the union of all
primary names and aliases for ACP-surfaced builtins — and use it as the
reserved set. Widen getRegisteredCommands parameter to ReadonlySet since
it only calls .has().
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Dispatch in AgentSession runs #tryExecuteExtensionCommand before
#tryExecuteCustomCommand, so the palette must reflect the same order.
Moving the extension-runner block before session.customCommands ensures
that on a name collision the advertised command matches what will
actually execute.
Update the test to assert the extension description wins over the
colliding custom TS description.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Update ordering comment in #buildAvailableCommands to document the
extension tier and explain why skills/custom TS commands intentionally
shadow extension commands (unlike interactive mode)
- Add CHANGELOG entry under [Unreleased]
- Add regression test: verifies extension commands surface in
available_commands_update and that a builtin-colliding extension
command is excluded via the reserved-set, with no duplicates
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
`#buildAvailableCommands` was omitting commands registered by extensions
via `extensionRunner.getRegisteredCommands()`. ACP clients (e.g. Zed)
never saw these in the `available_commands_update` notification, so they
couldn't forward the corresponding slash commands to the agent.
Mirrors the interactive-mode pattern: pass ACP builtin names as the
reserved set so extensions cannot shadow core commands.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
When the user presses Stop in Zed and immediately types a new message,
the new session/prompt RPC can arrive before (or without) a preceding
session/cancel notification. The previous guard threw an error in that
case, leaving the session stuck and blocking further interaction.
Replace the throw with an implicit cancel: call #beginCancelCleanup on
the unsettled turn so it resolves with stopReason:"cancelled", then let
#queuePrompt serialize the new prompt behind the abort cleanup as it
already does when session/cancel is called explicitly. #beginCancelCleanup
is idempotent so a concurrent explicit cancel notification is a no-op.
Updated the test to assert the new contract: overlapping prompt
auto-cancels the first turn and the second is processed normally.
Theo Mathieu