## What changed
Resolve `GuardianContextMode` once at session construction and share it across history retention, replay, evidence capture, compaction, and review consumers.
Extract synchronous reviewer context policy and consolidate Guardian V2 parent checkpoint selection so sampling and fast approval use the same eligibility checks. Preserve legacy checkpoint omission and strict compatibility requirements for thread-owned context.
## Testing
Adapt existing review and history tests to the explicit context mode. Move checkpoint tests alongside the extracted selector, retaining coverage for latest-item identity, unusable checkpoints, and complete-item size limits.
GitOrigin-RevId: c73fb6fea556921acfffe758a5f709d7494019dd
## Why
Queued prompts and `request_user_input` answers can be persisted in a different
order from when the host accepted them. Using persistence order can therefore
reorder retained instructions during replay or apply the wrong rollback
boundary.
## What changed
- Reserve an acceptance sequence for user prompts and elicitation responses
when retained thread context is enabled.
- Persist that sequence with user-message metadata and verified answers, and
use it when ordering, bounding, replaying, and rolling back retained context.
- Keep events without acceptance metadata compatible with legacy recording
order.
## Testing
- Cover delayed queued-input recording, rejected prompts, checkpoint replay,
compaction, rollback migration, and legacy retained-context behavior.
GitOrigin-RevId: 42d24465cc2033bc51c0c2da79d3c0884a8ad98a
## Why
When ordinary included usage is exhausted, eligible users need a way to keep working with the backend-provided Luna Reserve quota and return to their previous model once ordinary usage recovers.
## What changed
- Automatically switch eligible TUI tasks to Luna Reserve after an authoritative account usage read, preserving the prior model and reasoning effort per task.
- Restrict model selection while on Reserve to its supported reasoning levels, keep queued turns on the accepted model, and restore the saved model after a fresh read confirms recovery.
- Surface Reserve usage in the composer and `/status`, and expose `normalModelSlug` in app-server rate-limit snapshots so clients can use the associated model's display metadata without changing the request model.
- Poll account limits more frequently near exhaustion and retain compatibility with app servers that only accept the legacy null request parameters.
## Testing
- Added coverage for Reserve entry, queued-turn handling, task reconstruction, recovery authorization, account changes, model selection, usage rendering, and banner dismissal.
GitOrigin-RevId: 46a435e9531051f8cec3f6a46c4b9f7cfa479747
## Why
`request_permissions` paths and grants need to be evaluated against the selected executor environment, including its path convention, home directory, workspace roots, and temporary directories.
## What changed
- Resolve relative and home-relative permission paths using the executor context, reject mismatched path conventions and lossy paths, and support legacy `read` and `write` path lists.
- Move grant intersection into core so requested and granted permissions use the originating environment's sandbox context. Preserve deny entries conservatively when a special path cannot be resolved.
- Keep the full originating environment with pending permission requests so delayed responses are normalized against the same context.
## Testing
- Cover POSIX, Windows, UNC, relative, and home-relative path resolution and invalid path contexts.
- Verify end-to-end app-server grants are limited to the requested workspace scope and unresolved temporary-directory denies are preserved.
GitOrigin-RevId: 730a2aacd391262e92a6314f3a5b6c262e3dca10
## Why
Shell Snapshot V2 previously captured the login-shell environment lazily when a
command started, adding that work to the command path.
## What changed
- Start one asynchronous snapshot capture after turn hooks accept the turn, so
eligible local `exec_command` calls can reuse it without blocking the model.
- Use the turn's resolved tools, permissions, sandbox, environment, and shell
configuration, and skip prewarming for unsupported or network-managed cases.
- Cancel outstanding prewarming during shutdown. Keep failed speculative
captures retryable by real commands without consuming their retry budget.
## Testing
Added coverage for non-blocking capture and reuse, failure recovery, sandbox and
permission isolation, shutdown cancellation, hook-stopped turns, eligibility
gates, and concurrent capture retries.
GitOrigin-RevId: ff24a4f0328c06653e09d03fd0a3db1eccd65030
## Why
Git root discovery is optional metadata work, but filesystem probes can block. They should not exhaust Tokio's blocking pool, delay runtime shutdown, or prevent later turns from observing repository changes.
## What changed
- Add a shared `GitRootDiscovery` service that coalesces concurrent lookups for the same working directory and limits probes across directories.
- Run probes on detached threads, retain in-flight work across caller cancellation, and discard completed results instead of caching them.
- Use the service for turn and memory metadata enrichment, abort unused turn enrichment when its state is dropped, and limit memory metadata waits to one second.
## Testing
Add coverage for probe sharing, capacity limits, cancellation, fresh discovery, runtime shutdown, memory timeouts, and repositories restored after startup prewarming.
GitOrigin-RevId: bca46fc263e7a12a2f69146d8a0b3e7c7e0846cb
## Why
Resumed threads need to continue per-turn and per-thread token totals without scanning arbitrarily far beyond the latest compaction checkpoint.
## What changed
- Add durable `TokenUsageRecord` rollout items with response, turn, thread, session, and root-turn attribution.
- Restore accumulated usage on resume and snapshot the latest record plus the compaction response ID in compaction checkpoints.
- Preserve root-turn lineage in persisted turn context, while ensuring forked child threads start with their own usage totals.
## Testing
- Cover usage accumulation across multiple responses and resumed turns.
- Cover local and remote compaction checkpoints, invalid remote compaction output, rollout reconstruction, and fork isolation.
GitOrigin-RevId: ef9e0c4a9102a08a2c382be4cdac68c84353c90a
## Why
Compaction and host-injected context can rewrite the model-visible conversation
without changing what the user authorized. Treating those updates as authorization
changes prevents Guardian from reusing an otherwise valid review.
## What changed
- Track a host-owned user-message revision separately from the conversation history
generation.
- Advance the revision for genuine user messages and history resets, while preserving
it across compaction and internal context injection.
- Use message content-kind metadata to distinguish host context from user input,
conservatively treating unknown or incomplete metadata as user authorization.
## Testing
Added coverage that cached Guardian authorization survives compaction and internal
context, but is invalidated by user input and rollback.
GitOrigin-RevId: c7960fa182d1686042c23764d2abc58acbc4b882
## What changed
- Store the most recently captured execution `StepContext` in the active turn state.
- Keep speculative model-fallback captures from replacing that context, and update it only when the fallback is selected for remote compaction.
- Preserve the retained context when a turn completes or is interrupted.
## Testing
- Add coverage for completed and interrupted turns, successful primary compaction, fallback compaction, and interruption while a fallback is running.
GitOrigin-RevId: c7dee36ec8e88ba2f4fb4a9c0322e5c86a0caf0e
## What changed
- Preserve each contextual fragment's `ContentItemKind` when converting it to a response message.
- Classify user text, image, and audio content as `user.text`, `user.image`, and `user.audio` in their original order.
- Give additional context key-specific `additional_content.<key>` kinds and retain annotations on generated messages such as image resize notices and interrupted-turn markers.
## Testing
- Expand request-level coverage to verify content-kind annotations across mixed-modality turns, additional context, reminders, budgets, skills, model switching, subagents, and image resizing.
GitOrigin-RevId: 671e4545f39b335525c60d2a5df4608e8068cafc
## What changed
- Resolve user message admission when Core accepts the input as a new turn or a steer, without waiting for rollout persistence.
- Remove persistence- and hook-specific admission errors and task bookkeeping.
- Delete queued messages after Core accepts them for turn processing, including messages subsequently stopped by a prompt hook.
## Testing
- Update concurrent admission and queue service tests for acceptance-based admission.
GitOrigin-RevId: 7dcc311cfecd3d710733b420f2add546edb86153
## What changed
- Wrap response items with optional harness-owned metadata while keeping the persisted response payload shape backward compatible.
- Store compacted-history metadata in an aligned sidecar and reject malformed sidecars.
- Carry metadata through history normalization, compaction, resume, fork, truncation, migration, and rollout recording without sending it to model providers.
## Testing
- Cover legacy rollout compatibility, metadata serialization and validation, compaction retention, and resume/fork reconstruction.
GitOrigin-RevId: 17a6127c9048ff984cb4df4d97c0751a38c29f5f
## What changed
- Import skill metadata and mention helpers directly from `codex-skills`.
- Import host skill loading types directly from `codex-skills-extension`.
- Make the remaining `codex-core` skills module private and remove its skill API re-exports, including the `HostSkillsService` re-export from `codex-core-api`.
GitOrigin-RevId: 26bf8fffd980df9c4d0c308d7e983ceab9a5df82
## What changed
- Add lifecycle-backed gauges for in-flight and queued app requests, pending server requests and mailbox messages, active turns, and live MCP connections.
- Register each gauge on first use and decrement it when the tracked work or resource is dropped.
- Document the new `server/diagnostics` gauge names and account for the diagnostics request itself in the response example and test.
GitOrigin-RevId: 4bb82a3918af0b502d149b5d761ad669b2083f39
## Why
Persisted base instructions need to retain whether they were explicitly customized or generated from a model template. Without that distinction, forks and config-lock replays can treat model-generated instructions as custom and keep the wrong template after a model or personality change.
## What changed
- Record custom or model provenance with base instructions in rollouts and config locks.
- Preserve custom instructions across model changes, while allowing model-generated instructions to follow the selected model and personality.
- Keep legacy rollouts without provenance compatible and propagate provenance through forks and spawned agents.
## Testing
- Cover provenance serialization and legacy rollout decoding.
- Cover role personality changes, forked model changes, config-lock replay, and multi-agent configuration.
GitOrigin-RevId: 801c26559835f558027ce5112700c0cd31fcb21b
## Why
MCP tool handlers and their Code Mode definitions are immutable for the lifetime
of a stable MCP binding, so rebuilding them for every sampling step repeats schema
construction unnecessarily.
## What changed
- Cache MCP handlers per session and reuse them while the same binding remains
active, invalidating the cache when the binding changes.
- Lazily cache schema-free Code Mode definitions on MCP handlers while continuing
to rebuild definitions for step-local tools whose specifications may change.
- Reapply current app enablement, tool policy, exposure, and plugin schema budgets
whenever cached handlers are registered for a step.
## Testing
- Cover handler and Code Mode definition reuse, binding invalidation, changing app
policy, stable continuation schemas, and step-local custom tool updates.
GitOrigin-RevId: 69c5011fb97c08580ac476e7fb2286b1e92e0359
## Why
A model switch on the first turn has no prior turn settings to compare, and rolling that turn back could leave its model instructions in history.
## What changed
- Track the model that supplied the session's base instructions so first-turn switches inject the target model's instructions, including the selected personality.
- Preserve inherited custom base instructions when starting from an empty fork.
- Remove first-turn model-switch developer content during rollback so retries and cold resumes do not retain or duplicate stale instructions.
## Testing
Add integration coverage for first-turn switches, personality changes, empty-prefix forks, rollback retries, and cold resume.
GitOrigin-RevId: 9df532e9e337521ad3b315b77f8c98e039fb384e
## What changed
- Add a storage-neutral queue extension for listing, adding, editing, reordering, deleting, and explicitly starting queued user messages.
- Dispatch queued messages in FIFO order when a thread becomes idle after a completed turn, while leaving them queued after interruptions or failures.
- Wait for user input to be persisted before removing its queue item, and report hook rejection, task termination, and persistence failures distinctly.
- Snapshot local image and audio attachments into validated data URLs before storing them, enforce input limits, and emit `ThreadQueueChanged` events after mutations.
## Testing
- Cover queue mutations, automatic and explicit dispatch, interruption and failure behavior, hook rejection, invalid records, input limits, and local attachment snapshotting.
GitOrigin-RevId: 14f39d7c8cbaccb767a6da84b82cced0edfcb4b9
## What changed
- Move host skill discovery, snapshot caching, and configuration handling from
`codex-core-skills` to `codex-skills-extension`.
- Rename `SkillsService` and `SkillsLoadInput` to `HostSkillsService` and
`HostSkillsLoadInput` to distinguish the host implementation from other skill
providers.
- Keep shared loading primitives and skill outcome modeling in
`codex-core-skills`.
GitOrigin-RevId: d81a21791d2ef8c066e157e7e538b8cb7ee4c24b
## Why
App-server clients need to advertise structured MCP extension settings, including supported MCP App UI MIME types, rather than only opting into OpenAI form elicitation with a boolean.
## What changed
- Add an `extensions` map to initialize capabilities and preserve `mcpServerOpenaiFormElicitation` as a legacy alias for `openai/form`.
- Capture the declared extension profile when a thread is started, resumed, or forked, propagate it to subagents, and advertise it to downstream MCP servers during initialization.
- Keep the profile stable for the lifetime of the loaded session instead of changing it on later turns or direct tool calls.
## Testing
- Cover extension conversion, downstream MCP initialization, session isolation, legacy form support, and subagent inheritance.
GitOrigin-RevId: fbcedbb74ce788e574b0f884a4c45c4cedb9de54
## What changed
- Recognize the `codex_strict_auto_review` MCP elicitation marker and route
marked approval requests through the configured automatic reviewer.
- Accept only canonical automatic-review approvals and fail closed without a
user prompt when review is unavailable, denied, malformed, disallowed by
policy, or requests persistence.
- Validate Codex Apps approval metadata against the active MCP invocation and
build the review request from the trusted invocation details.
## Testing
- Add unit coverage for strict review decisions, malformed markers, reviewer
failures, explicit policy denials, and lifecycle cleanup.
- Add app-server round-trip coverage for approvals, denials, configuration
constraints, spoofed metadata, persistent requests, and subsequent turns.
GitOrigin-RevId: af73b45a9b63118f25f6429d95b3222c5dbb59fb
## What changed
- Replace `SessionTaskContext` with `Arc<Session>` in the session task run and abort interfaces.
- Read extension data from `TurnContext` during regular turns instead of storing and forwarding a duplicate reference through the running task.
GitOrigin-RevId: 04bc0d324e5ba781f35ac9c61e811888208b74ae
## What changed
- Add the opt-in `executed_tool_call_metadata` feature to record model-attempted direct and code-mode tool calls.
- Attach recorded names and arguments to the matching tool output in the next Responses request, preserving the metadata across sampling retries.
- Bound pending calls and argument sizes, and emit truncation metadata when limits are exceeded.
## Testing
- Cover disabled-by-default behavior, namespaced and nested calls, blocked or failed attempts, retry handling, and recorder limits.
GitOrigin-RevId: 5ce917b73e797b0f6904d46477b9d2cf8ea71bf2
## What changed
- Store a request-logging-disabled `RouteAwareClientPool` in session services and use it for file creation, blob upload, and finalization requests.
- Preserve the existing system-root fallback for transport-default proxy routes, and emit a warning event when that fallback is used.
- Remove the direct `reqwest` dependency from `codex-api` now that uploads use the shared HTTP abstraction.
## Testing
- Add coverage for completing multiple uploads through a shared client pool.
GitOrigin-RevId: 01842415cad2d349a6bd3869abb3332e8110ebce
## What changed
- Make `McpRuntime` own the published MCP configuration, connections, elicitation routing, and selected capability roots for a thread.
- Capture immutable MCP bindings for model steps and tool calls so in-flight work keeps a consistent connection set and approval authority while refreshed state is published atomically.
- Mark MCP state dirty when relevant configuration, plugins, environments, authentication, or elicitation capabilities change, then rebuild it before the next sampling request or out-of-band MCP operation.
- Separate config reloads from server invalidation: `ReloadMcpConfig` applies resolved MCP inputs, while `RefreshMcpServers` requests reinitialization from the thread's latest state.
## Testing
- Cover refreshed state visibility for existing turns, stable step bindings, current approval authority, cancelled refresh retries, resource-client reconciliation, and Apps recovery between sampling requests.
GitOrigin-RevId: 59eabb1aa8dc083426bd18ef4d3630508f376401
## What changed
- Use `McpConnectionSet` as the primary name for the collection of active MCP
server connections throughout `codex-mcp` and its callers.
- Keep `McpConnectionManager` as a backward-compatible type alias.
GitOrigin-RevId: 9753838b4fd0a0b6429a9c4af0c077137616f388
## What changed
- Change `ReviewDecision::Denied` to carry a rejection string and update the generated protocol schemas.
- Preserve specific rejection reasons through command, patch, network, MCP, delegated, and automatic approval flows so tool results can return them to the model.
- Distinguish invalid approval responses from user declines and truncate oversized rejection messages before adding them to model context.
## Testing
- Add coverage for denial serialization, invalid app-server approval responses, dropped automatic reviews, network rejection propagation, and rejection-message truncation.
GitOrigin-RevId: d48a55643aa03c7ecfa08a8e4b669913eec5464a
## What changed
- Add a thread-owned `McpRuntime` that atomically publishes connection-set
replacements and owns shutdown.
- Route `McpResourceClient` snapshots and cache keys through that runtime so
extension clients observe refreshed MCP servers.
- Keep model-step configuration snapshots separate from ownership of live MCP
connections.
## Testing
- Add coverage that captures an extension's session resource client, refreshes
the MCP server configuration, and verifies that the existing client sees the
newly published server.
GitOrigin-RevId: 351b2e25594b8eb6b91e50ff8d06b7d49768ae1c
## What changed
- Let deferred environments provide selected capability roots with their ready signal.
- Validate that those roots have unique, non-empty IDs, belong to the registering environment, and stay within the root limit.
- Include roots from ready turn environments when resolving MCP contributions, and refresh the MCP runtime when the selected root set changes.
- Expose the exact ready root set to MCP contributors so executor plugins become available with their environment.
## Testing
- Cover ready-root propagation, validation failures, replacement isolation, reconnection, and MCP plugin availability refresh.
GitOrigin-RevId: ec3498aab1164824025094e96a9b1063b7b731ad
## Why
Token-budget sessions need a final opportunity to preserve important state before the current context is compacted.
## What changed
- When `auto_compact_fallback_prompt` is configured, reserve `auto_compact_fallback_buffer_tokens` beyond the base auto-compaction limit.
- At the base limit, inject the fallback prompt once as a developer message while keeping the normal tool surface available. Continue reporting zero base-window tokens during this reserve.
- Roll over when the reserve or full model context is exhausted, and skip the fallback when a new context was already requested.
## Testing
Added coverage for using the fallback reserve, rolling over after it is exhausted, and bypassing the fallback for an explicit `new_context` request.
GitOrigin-RevId: 32983ecadc5cb4927e15ecd7b1e896cb52a9ad3b
## Why
Sparse and out-of-order rate-limit updates can omit account metadata or let an
older full read overwrite a newer workspace hard stop. Credit availability also
does not always include a displayable balance.
## What changed
- Propagate the backend's `spendControlReached` state through rate-limit
snapshots and the app-server protocol, preserving it across sparse updates.
- Keep workspace hard-stop classifications for error routing and ignore stale
full-read results after a newer hard-stop notification.
- Base TUI warnings and model-switch prompts on credit availability and hard-stop
state, and show `Credits: Available` when a finite balance is hidden.
## Testing
Added coverage for spend-control mapping, snapshot merging, out-of-order reads,
workspace error routing, warning and prompt behavior, and credit status display.
GitOrigin-RevId: 18b09d493d5a620a7c079fb721ec50b38dd5b04c
## Why
Code-mode tool results could return to the model while an MCP
elicitation was still waiting for user input. This differed from
parallel tool calling and could let the model continue before the user
resolved the request.
We need one session-level view of outstanding elicitations so tool
runtimes can consistently hold results until every pending elicitation
is resolved.
## What changed
- Added a counted, session-owned ElicitationService with RAII
registrations.
- Registered both core-originated and server-originated MCP elicitations
with the service.
- Migrated out-of-band elicitation tracking and unified exec timeout
pausing to the shared service.
- Made code-mode functions.exec and functions.wait capture their runtime
result normally, then hold it before returning while an elicitation is
outstanding.
- Kept terminate: true immediate; only its result is held.
- Preserved model-visible wall time across the elicitation hold.
- Kept the behavior session-scoped, with concurrent elicitations holding
the pause until all registrations are released.
## Why
Generic plugin guidance is currently emitted only with initial context
from host plugin state. An executor-selected plugin can become available
later in the same turn, making its skills and tools usable without ever
telling the model how plugin capabilities should be used.
## What
- project every ready selected plugin package, including skill-only
plugins
- carry plugin availability with the exact MCP runtime projection while
preserving MCP manager reuse when servers and connectors are unchanged
- move generic plugin guidance from the static initial-context path into
persisted World State
- recognize legacy and retained plugin fragments so resume and
compaction do not duplicate guidance
## Testing
- `just test -p codex-mcp-extension`
- `just test -p codex-core plugins_instructions`
- `just test -p codex-core
plugin_availability_change_reuses_the_mcp_manager`
- `just test -p codex-app-server --test all selected_capabilit`
## Why
Selected plugin metadata is stable, but MCP processes are live runtime
state. They need different lifetimes:
- the MCP extension caches manifest, MCP, and connector declarations for
each stable selected root;
- each model step projects that cached metadata through the roots that
resolved as ready for that exact step;
- the MCP manager is rebuilt only when that availability projection
changes.
This matches executor skills: both features consume the same resolved
step roots instead of inferring readiness from the turn's selected
environments.
## Behavior
```text
E1 not ready for this step
-> no E1 MCP servers or connectors
-> cached plugin metadata stays in ext/mcp
E1 becomes ready
-> reuse cached metadata
-> publish one MCP runtime containing E1 capabilities
same ready roots on the next step
-> reuse the exact runtime; no rediscovery and no MCP restart
resume
-> create new extension thread state and a new MCP runtime
```
All model-facing consumers use the same step snapshot:
```text
resolved selected roots
|
v
extension MCP/connector projection
|
v
{ MCP config, connector snapshot, MCP manager }
|
+-> advertise model tools
+-> build app/connector tools
+-> execute MCP calls
```
## Cache contract
The existing MCP extension owns a cache keyed by the full
`SelectedCapabilityRoot`:
```rust
let state = thread_store.get_or_init(SelectedExecutorPluginMcpState::default);
```
The cache lives with extension thread state. Environment availability
filters projection but does not invalidate metadata. Resume creates new
thread state. There is no file watcher or executor generation because
contents behind a stable environment/root are assumed stable.
## What changes
- Keeps executor plugin discovery and cached metadata in `ext/mcp`.
- Caches MCP and connector declarations together per selected root.
- Uses the step's already-resolved capability roots, including lazy
environments that are not turn environments.
- Reuses the current MCP runtime when the ready-root projection is
unchanged.
- Uses the same step MCP manager and connector snapshot for
model-visible tools and execution.
- Resolves direct thread-scoped MCP requests from the current
selected-root projection.
## Deliberately out of scope
- `app/list` remains based on the latest global host-plugin state; this
PR does not make its response or notifications thread-specific.
- `required = true` startup semantics do not apply to delayed executor
MCP activation.
- No filesystem/content invalidation.
- No transport-disconnect watcher.
- No executor generations or environment replacement semantics.
- No client sharing across complete manager replacements.
## Stack
1. Extension-owned World State sections.
2. Project executor skills through World State.
3. Pin one MCP runtime to each model step.
4. **This PR:** project selected MCP and connector state from
extension-owned metadata.
5. Integration coverage for selected capability availability and resume.
## Verification
-
`selected_plugin_servers_use_managed_requirements_for_the_selected_root_id`
- The stacked integration PR covers unavailable to ready activation,
unchanged-runtime reuse, skills, MCP tools, connector attribution, and
cold resume.
## Why
An MCP refresh can replace the session's current manager while a model
step is still running. The step must execute calls through the same
manager whose tools it advertised.
## Boundary
```text
current session MCP runtime
|
| capture once for this model step
v
StepContext.mcp
- exact MCP config
- exact connection manager
- exact runtime environment context
```
```rust
pub struct McpRuntimeSnapshot {
config: Arc<McpConfig>,
manager: Arc<McpConnectionManager>,
runtime_context: McpRuntimeContext,
}
```
## Example
```text
step A captures runtime A and advertises A's tools
refresh publishes runtime B
step A tool call -> runtime A
next step -> runtime B
```
Capturing the snapshot is only an `Arc` clone. It does not restart MCPs
or make an RPC.
## What changes
- Captures one MCP runtime in `StepContext`.
- Uses it for tool planning, tool calls, resources, approvals, connector
attribution, and elicitation.
- Publishes replacement runtimes atomically.
- Lets an old runtime live only while an in-flight step or request still
holds its `Arc`.
Most of this diff is mechanical routing from the session-global manager
to `step_context.mcp`; it does not introduce selected-plugin discovery
yet.
## What does not change
- No plugin or extension migration.
- No new MCP cache policy.
- No environment file watching.
- No client sharing between separate managers.
## Stack
1. Extension-owned World State sections.
2. Project executor skills through World State.
3. **This PR:** pin one MCP runtime to each model step.
4. Project selected MCP/app/connector metadata by environment
availability.
5. One end-to-end integration scenario.
## Why
`selectedCapabilityRoots` is durable thread intent: “use this capability
root from environment `worker`.”
The important product assumption is:
> One environment ID always names the same logical executor and stable
contents.
`worker` does not silently change from executor A to an unrelated
executor B. The process-local connection handle for `worker` can still
be replaced while Codex is running, though, for example when
`environment/add` registers a fresh handle for the same logical
environment.
The thread should persist only the stable selection. Each model step
should pair that selection with the exact ready handle captured for that
step.
## The boundary
```text
persisted thread intent
plugin@1 -> environment "worker"
|
| capture the current step
v
model-step view
unavailable, or
plugin@1 + worker's exact captured ready handle
```
The environment ID is the stable identity and cache key. The
`Arc<Environment>` is only a process-local handle retained so consumers
of one model step use the same captured environment. It is never
persisted and it does not imply different environment contents.
## What changes
### Persist the stable selection
Selected roots are written into `SessionMeta` and restored with the
thread. Forked subagents inherit the same selections, including
bounded-history forks.
Only stable data is persisted: root ID, environment ID, and root path.
### Capture readiness together with the exact handle
The environment snapshot records:
```rust
environment_id -> Some(Arc<Environment>) // ready in this step
environment_id -> None // still starting in this step
```
This prevents readiness and execution from coming from different
registry snapshots.
For example:
```text
step snapshot: worker -> handle A, ready
environment/add: worker -> fresh handle B for the same logical environment
current step: plugin@1 still uses captured handle A
```
Without carrying handle A in the snapshot, the resolver could combine “A
was ready” with handle B and treat B as ready before it had finished
starting.
This does not change cache invalidation. Stable capability metadata
remains identified by environment ID and capability root. Replacing a
process-local handle under the same stable environment ID does not
invalidate or rediscover that metadata.
### Resolve availability per model step
- A ready captured environment produces resolved roots using its
captured handle.
- A starting, missing, or failed environment is omitted from that step.
- A selected lazy environment that is outside the turn's captured
environment set is asked to start, and a later step can observe it as
ready.
- No capability files are scanned here.
Transient transport disconnects remain the remote client's reconnect
concern. This PR models initial attachment/readiness; it does not add
live socket-connectivity state.
## Example
```text
thread selection: plugin@1 -> environment "worker"
step 1: worker is starting -> plugin@1 unavailable
step 2: worker is ready -> plugin@1 resolves through worker's captured handle
step 3: fresh local handle -> current step remains pinned; a later step captures its own view
```
Temporary unavailability does not discard the durable selection. Later
PRs can retain stable metadata caches while projecting only currently
available capabilities into model-visible World State.
## Compatibility
The app-server request shape does not change. Older rollouts without
`selected_capability_roots` deserialize to an empty list.
## Stack
1. **This PR:** persist stable selected roots and resolve them through
an exact model-step handle.
2. #29960: cache stable skill metadata and project available skills into
World State.
3. #29946: cache stable plugin declarations and manage the separate live
MCP runtime.
## Why
With deferred executors, a turn can begin before a remote environment
attaches. AGENTS.md discovery previously ran only during session setup,
so instructions from a later environment never reached the model or the
session instruction sources.
WorldState persistence has now landed, so this uses the durable
model-visible baseline directly instead of carrying a temporary
resume/fork compatibility path.
## What
- Add an `AgentsMdManager` in `SessionServices` to own host
instructions, loaded state, and refresh caching.
- When `DeferredExecutor` is enabled, refresh AGENTS.md when attached
environment selections change and freeze the result in the corresponding
`StepContext`.
- Represent AGENTS.md as a persisted WorldState section for every
session, with bounded initial, replacement, and removal updates.
- Remove duplicate AGENTS.md state and rendering from
`SessionConfiguration` and `TurnContext`.
- Build initial context, per-request updates, and compaction context
from the same step-scoped value.
- On resume and fork, compare current instructions with the restored
WorldState baseline and inject a replacement exactly once when they
differ.
Builds on #29833, #29835, and #29837.
## Tests
- Covers a remote environment becoming ready mid-turn, with AGENTS.md
appearing on the next request exactly once and updating canonical
instruction sources.
- Covers full, unchanged, replaced, and removed AGENTS.md WorldState
rendering.
- Covers changed instructions across cold resume and fork without
duplicate reinjection.
- Covers remote-v2 compaction retaining creation-time instructions in
the live session and cold resume appending one replacement when the
source changed.
- Ran focused `codex-core` AGENTS.md, WorldState, and context-update
test suites.
## Why
When `Feature::TokenBudget` is enabled, compaction should behave like
`new_context`: start a fresh context window with the standard injected
context, without asking the server to summarize old history and without
carrying prior user or assistant messages into the next model request.
This is still a compaction operation from the client lifecycle
perspective. Manual `/compact` and auto-compaction should keep the same
observable side effects that clients and hooks expect, including compact
hooks and `TurnItem::ContextCompaction`.
## What changed
- Added `compact_token_budget` to run token-budget manual and inline
auto-compaction through a shared compaction lifecycle.
- Split pending `new_context` requests from forced context-window
startup: `take_new_context_window_request()` consumes pending requests,
and `start_new_context_window()` installs a fresh context window.
- Routed token-budget manual `/compact` and inline auto-compaction to
install a fresh context window locally instead of calling server/local
summarization.
- Preserved compact lifecycle side effects for token-budget compaction
by running pre/post compact hooks and emitting `ContextCompaction` item
start/completion events.
- Updated token-budget tests to assert fresh window IDs, absence of
server-side compaction calls, dropped prior transcript messages/tool
output after reset, and compact hook/item lifecycle behavior.
## Testing
- `just test -p codex-core
token_budget_context_uses_new_window_after_compaction`
- `just test -p codex-core token_budget_compaction_runs_compact_hooks`
- `just test -p codex-core
token_budget_mid_turn_auto_compaction_resets_before_active_follow_up`
---------
Co-authored-by: pakrym-oai <pakrym@openai.com>
## Why
PR #29494 made context-window IDs visible to the model by wrapping the
token-budget window payload in `<context_window>`, but rollout JSONL
consumers still could not see the initial window identity by tailing the
session file. Compacted rollout items carry window IDs only after
compaction has happened, so a session with no compaction had no durable
JSONL record for window 0.
This change gives tailing consumers a stable initial-window record at
session creation time.
## What Changed
- Added `session_meta.context_window.window_id` for the initial
context-window identity.
- `CreateThreadParams` now requires `initial_window_id: String`, so
thread-store callers cannot accidentally create new threads without
window-0 metadata.
- Live thread creation derives the persisted initial window ID from the
same `AutoCompactWindowIds` used to initialize `SessionState`, keeping
runtime state and JSONL metadata aligned.
- Rollout reconstruction uses `session_meta.context_window.window_id` as
the initial-window fallback and derives `window_number = 0`,
`first_window_id = window_id`, and `previous_window_id = None`
internally.
- Fork reconstruction intentionally uses the same rollout reconstruction
path; consumers that need to distinguish copied initial-window metadata
can use the rollout `thread_id`.
- Legacy compactions without `window_number` still use compaction-count
fallback accounting instead of being reset to window 0 by the
initial-window fallback.
- Compacted rollout metadata still takes precedence once compaction
records exist, preserving the richer chain fields there.
## JSONL Shape
Real rollout JSONL is one object per line. This example is expanded for
readability, but shows the new initial `session_meta.context_window`
record followed by the existing compacted rollout item shape that also
carries window IDs:
```jsonl
{
"timestamp": "2026-06-22T12:00:00.000Z",
"type": "session_meta",
"payload": {
"session_id": "<THREAD_ID>",
"id": "<THREAD_ID>",
"timestamp": "2026-06-22T12:00:00.000Z",
"cwd": "/repo",
"originator": "codex",
"cli_version": "0.0.0",
"source": "cli",
"model_provider": "<MODEL_PROVIDER>",
"context_window": {
"window_id": "<INITIAL_WINDOW_ID>"
}
}
}
...
{
"timestamp": "2026-06-22T12:34:56.000Z",
"type": "compacted",
"payload": {
"message": "<COMPACTION_SUMMARY>",
"replacement_history": [
"..."
],
"window_number": 1,
"first_window_id": "<INITIAL_WINDOW_ID>",
"previous_window_id": "<INITIAL_WINDOW_ID>",
"window_id": "<NEXT_WINDOW_ID>"
}
}
```
The nested `context_window` object is intentional: it gives rollout
consumers a stable namespace for context-window metadata while only
writing the non-derivable initial `window_id`. For the initial window,
`window_number`, `first_window_id`, and `previous_window_id` are derived
internally instead of being written to the rollout.
## Verification
- `just test -p codex-protocol`
- `just test -p codex-rollout
recorder_materializes_on_flush_with_pending_items`
- `just test -p codex-core reconstruct_history`
- `just test -p codex-core
record_initial_history_reconstructs_forked_transcript`
- `just test -p codex-thread-store`
- `just test -p codex-state`
- `just test -p codex-app-server
thread_read_returns_summary_without_turns`
- `just test -p codex-rollout persistence_metrics`
## Why
The token-budget feature reports coarse remaining-context milestones,
but it does not give the model a configurable wrap-up prompt before
automatic compaction. A strict threshold-crossing check can also miss
resumed or reconfigured windows that are already inside the threshold.
## What changed
- Add structured `[features.token_budget]` configuration for an absolute
`reminder_threshold_tokens` and bounded `reminder_message_template`;
`{n_remaining}` is expanded when the reminder is delivered.
- Compute remaining tokens against the next effective auto-compaction
boundary, including scoped `body_after_prefix` accounting and the full
context-window limit.
- Make reminder delivery level-triggered before and after sampling, with
one-shot state owned by `AutoCompactWindow` and re-armed on compaction,
`new_context`, restore, or history replacement.
- Leave the existing initial full-window token-budget context, 25/50/75%
notices, and token-budget tools unchanged.
- Persist the resolved feature configuration in the session config lock
and regenerate the config schema.
## Validation
- `just test -p codex-core token_budget`
- `just test -p codex-core
token_budget_reminder_emits_after_crossing_compaction_threshold`
- `just test -p codex-core auto_compact_window`
- `just test -p codex-core
lock_contains_prompts_and_materializes_features`
- `just test -p codex-features`
- `just test -p codex-config`
## Why
The rendered `<token_budget>` fragment identifies the thread and current
context window, but it does not expose enough lineage to identify the
first window in the thread or the immediately preceding window. Those
IDs also need to remain stable across compaction, resume, and rollback.
## What changed
- Track first, previous, and current UUIDv7 context-window IDs in
auto-compaction state.
- Render `thread_id`, `first_window_id`, `previous_window_id`, and the
current window ID in the full `<token_budget>` fragment.
- Persist the first and previous window IDs in compacted rollout
checkpoints and restore them during rollout reconstruction.
- Preserve compatibility with older compacted records that do not
contain the new optional fields.
- Update focused state, rendering, reconstruction, rollback, and
serialization coverage.
## Validation
- `just test -p codex-core token_budget`
- `just test -p codex-protocol compacted_item::tests`
- `just test -p codex-core tracks_prefill_and_window_boundaries`
- `just test -p codex-core
reconstruct_history_uses_replacement_history_verbatim`
- `just test -p codex-core
thread_rollback_restores_cleared_reference_context_item_after_compaction`
## Why
The token-budget context currently identifies a context window by its
thread-local sequence number. A UUIDv7 gives the model a stable opaque
identity that remains fixed for a window and rotates when compaction or
`new_context` starts the next one.
## What changed
- Preserve the existing monotonic value as `window_number` and add a
UUIDv7 `window_id` to `CompactedItem`.
- Generate and rotate the UUID with auto-compaction window state,
persist it alongside the number, and reconstruct it on resume and
rollback.
- Accept legacy compacted rollout records where the numeric `window_id`
represented the window number.
- Use the UUID only in token-budget context; existing request headers
and metadata continue using `thread_id:window_number`.
## Testing
- `just test -p codex-protocol compacted_item::tests`
- `just test -p codex-core token_budget`
Stacked on #28822.
## Summary
- add a host-injectable current-time provider with a built-in system
implementation
- record UTC developer reminders in history immediately before due model
requests
- keep cadence state per session and force a refresh after compaction
This does NOT include the app server client <-> server clock logic. This
PR is only for the reminder message & system clock that will be used in
prod.
## Testing
- `just test -p codex-core varlatency_`
- `just clippy -p codex-core -p codex-app-server -p codex-mcp-server -p
codex-thread-manager-sample`
- `just fmt`
## Why
Host skill discovery was still exposed as a manager even though it is a
process-owned service shared by sessions, the app-server catalog, and
file-watcher invalidation. The skills extension also consumed an ad hoc
loaded-skills wrapper instead of a named immutable snapshot.
## What changed
- replace `SkillsManager` with concrete `SkillsService`
- make the service cache and return immutable `HostSkillsSnapshot`
values
- migrate the skills extension host provider to the snapshot boundary
- migrate app-server catalog, watcher, and invalidation paths to the
service
This keeps the service limited to host discovery, caching, roots, and
invalidation. Catalog rendering and invocation remain extension
responsibilities for the next stacked change.
## Why
Shell snapshots are currently session-scoped even though shell and cwd
are properties of a selected turn environment. That makes snapshot
refresh depend on separate session-cwd plumbing, prevents retained
environments from retaining their snapshot work, and can make snapshot
construction use a different shell than command execution.
This follows #27955 by making the retained thread-environment service
own environment snapshot lifecycles. Session configuration remains the
requested selection state, while `ThreadEnvironments` remains the source
of successfully resolved environments.
## What changed
- Configure the shell-snapshot builder before initial environment
resolution.
- Start each local environment snapshot task when its `TurnEnvironment`
is built and retain that shared task while environment ID and cwd still
match.
- Inherit retained environment snapshots into spawned child threads.
- Carry the selected `TurnEnvironment` through shell runtimes so
snapshot construction and command execution use the same
environment-specific shell and cwd.
- Load project instructions and warm plugins/skills after initial
environment resolution.
- Continue decoding invalid UTF-8 instruction files lossily without
emitting a startup warning.
- Keep requested selections in `SessionConfiguration`; failed or
duplicate resolutions only affect the resolved environment snapshot.
## Validation
- `cargo check -p codex-core --tests`
- `just test -p codex-home instructions` (6 passed)
- Focused environment, instruction, shell-snapshot, and user-shell tests
(84 passed)
- Focused shell-snapshot, user-shell, and unified-exec tests (126
passed; two event-timing tests passed on retry)
## Why
Selected execution environments are thread-scoped resources, but startup
and turn construction repeatedly resolved their IDs and working
directories. That discarded existing environment handles and shell
metadata even when a selection had not changed.
Session configuration updates also need to affect future turns without
changing the resolved environment set already captured by a running
turn.
## What changed
- Create a `ThreadEnvironments` service inside `Codex` from the spawned
`EnvironmentManager` and raw environment selections, then store it on
`SessionServices`.
- Split service construction from `update_selections`, allowing session
configuration updates to mutate the resolved set in place.
- Retain an existing `TurnEnvironment` when its environment ID and
working directory match; resolve only added or changed selections and
remove selections that are no longer present.
- Normalize duplicate IDs by keeping the first selection and skip
individual selections that fail to resolve instead of rejecting the
entire update.
- Give each `TurnContext` a cloned `TurnEnvironmentSnapshot`, so later
session configuration updates affect future turns without rewriting an
active turn.
- Reuse the service-owned environment manager and resolved snapshot for
startup work, MCP initialization, and child-thread spawning instead of
flowing resolved environments through spawn arguments.
## Test plan
- `cargo check -p codex-core --tests`
- `just test -p codex-core environment_selection`
- `just test -p codex-core turn_environments`
- `just test -p codex-core
session_update_settings_does_not_rewrite_sticky_environment_cwds`
- `just test -p codex-core
default_turn_does_not_overlay_legacy_fallback_cwd_onto_stored_thread_environments`
## Why
Tool router construction rebuilds the deferred-tool BM25 index during
session initialization and before each sampling continuation, even when
the searchable tool metadata is unchanged. Local profiling measured
`append_tool_search_executor` at roughly 113 ms per continuation, making
repeated index construction the largest measured router-building cost.
## What changed
- Add a session-scoped `ToolSearchHandlerCache` so continuations and
user turns can reuse the existing handler.
- Key reuse on the complete ordered `Vec<ToolSearchInfo>`, rebuilding
when searchable text, loadable tool specs, source metadata, or ordering
changes.
- Build handlers outside the cache lock and recheck before publishing
them, avoiding holding the mutex during index construction.
## Verification
- `cache_reuses_identical_search_infos_and_rebuilds_changed_inputs`
covers exact cache reuse and invalidation when the ordered search
metadata changes.
- Local rollout profiling showed the initial router build populating the
cache and unchanged later continuations reusing it:
- uncached: 118 ms median across 14 spans from 3 rollouts
- cached: 4 ms median across 12 spans from 3 rollouts
## Why
Shell snapshot lifecycle state was split between `Shell` and
`SessionServices`: `Shell` carried the receiver while session code
exposed and forwarded the raw sender. That coupled shell identity to
mutable snapshot state and made refresh, inheritance, and file lifetime
harder to reason about.
## What changed
- make each `Arc<ShellSnapshot>` represent one cwd-specific snapshot
generation
- store the active generation in `SessionServices` with `ArcSwapOption`
- have construction start the background build and expose only a
cwd-validated snapshot path
- use `ShellSnapshotFile` ownership to delete snapshot files
automatically
- pass snapshot paths explicitly to shell runtimes instead of storing
snapshot state on `Shell`
- preserve inherited and in-flight generations by pinning their `Arc`
while they are in use
## Test plan
- `cargo check -p codex-core --lib`
- `just test -p codex-core 'shell_snapshot::tests'`
- `just test -p codex-core
shell_command_snapshot_still_intercepts_apply_patch`
- `just test -p codex-core
shell_snapshot_deleted_after_shutdown_with_skills`
## Why
`selectedCapabilityRoots` belongs to one thread, but MCP contributors
previously received only the global Codex config. That left no clean way
for a selected executor capability to contribute MCP servers to its own
thread.
## What this PR does
- Gives MCP contributors a small context containing the config and, for
a running thread, its frozen host-seeded inputs.
- Uses the same thread inputs during startup, status queries, refreshes,
and skill dependency checks.
- Keeps threadless MCP operations and the existing hosted Apps behavior
unchanged.
- Adds coverage showing that two threads resolve independent
registrations and that later lifecycle mutations do not change the
frozen MCP inputs.
This PR does not discover plugin manifests, add MCP servers, or launch
anything new. It only establishes the thread-scoped registration
boundary.
## Follow-ups
- Resolve selected executor plugin roots through their owning
environment filesystem.
- Convert their stdio MCP declarations into environment-bound
registrations and add an executor MCP end-to-end test.
## Verification
- `just fmt`
- `cargo check --tests -p codex-protocol -p codex-extension-api -p
codex-mcp-extension -p codex-core -p codex-app-server`
Tests and Clippy were not run.
felixxia-oai