deepagents-code 0.1.86__tar.gz → 0.1.88__tar.gz

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/CHANGELOG.md

@@ -2,6 +2,29 @@
 
 # Deep Agents Code Changelog
 
+## [0.1.88](https://github.com/langchain-ai/deepagents/compare/deepagents-code==0.1.87...deepagents-code==0.1.88) (2026-10-22)
+
+### Bug Fixes
+
+- Fixed async subagents to pass through `webhook` configuration.
+- Fixed bridge compatibility with Python 3.14.
+- Silenced no-op automatic cache offloads.
+
+## [0.1.87](https://github.com/langchain-ai/deepagents/compare/deepagents-code==0.1.86...deepagents-code==0.1.87) (2026-10-20)
+
+### Features
+
+- Resume threads in non-interactive mode ([#7011](https://github.com/langchain-ai/deepagents/pull/7011)).
+- Compact large idle sessions before the prompt cache expires ([#6956](https://github.com/langchain-ai/deepagents/pull/6956)) and show the prompt-cache expiration time ([#7005](https://github.com/langchain-ai/deepagents/pull/7005)).
+- Support custom Slack OAuth app identity ([#7061](https://github.com/langchain-ai/deepagents/pull/7061)).
+- Load older threads from the thread switcher ([#7051](https://github.com/langchain-ai/deepagents/pull/7051)).
+
+### Bug Fixes
+
+- Stop the local server when its owner dies ([#7038](https://github.com/langchain-ai/deepagents/pull/7038)).
+- Support non-ASCII file names in `@` mentions ([#7023](https://github.com/langchain-ai/deepagents/pull/7023)).
+- Correct misleading `/reload` change summaries ([#7063](https://github.com/langchain-ai/deepagents/pull/7063)).
+
 ## [0.1.86](https://github.com/langchain-ai/deepagents/compare/deepagents-code==0.1.85...deepagents-code==0.1.86) (2026-10-16)
 
 ### Features

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/DEVELOPMENT.md

@@ -69,6 +69,30 @@ make lint
 
 Run `make help` to see every available target.
 
+## Custom Slack MCP app identity
+
+Slack's hosted MCP server attributes sent messages to the OAuth app used to connect. To use your own registered Slack app instead of Deep Agents Code, add an `oauth` object to its `.mcp.json` server entry:
+
+```json
+{
+  "mcpServers": {
+    "slack": {
+      "type": "http",
+      "url": "https://mcp.slack.com/mcp",
+      "auth": "oauth",
+      "oauth": {
+        "client_id": "${SLACK_CLIENT_ID}",
+        "client_secret": "${SLACK_CLIENT_SECRET}"
+      }
+    }
+  }
+}
+```
+
+Enable MCP access for your Slack app and register `http://localhost:3118/callback` as its redirect URL. Set the environment variables and run `dcode mcp login slack` to authorize it. Confidential apps use `client_secret_post`; omit `client_secret` only for an app registered as a public client. Keep secrets in environment variables, not committed config. The registration, including any secret, is saved in the existing restricted local token store.
+
+Credentials are isolated by server name, URL, and configured client ID, so changing apps requires a separate login and never reuses the default app's grant. Re-run login after rotating a client secret. Omitting `oauth` preserves the existing app and credentials. This selects the authenticated app; it does not let a message supply arbitrary attribution or suppress Slack's footer. These settings apply to dcode; Talon's separate OAuth implementation is unchanged.
+
 ## LangSmith tracing projects
 
 When LangSmith tracing is enabled and a `LANGSMITH_API_KEY` is available, `deepagents-code` traces its own agent runs to LangSmith. A key alone is not enough for environment-supplied keys — set `LANGSMITH_TRACING=true`. Keys stored via `/auth` auto- enable tracing unless you explicitly opt out (for example `DEEPAGENTS_CODE_LANGSMITH_TRACING=false`).
@@ -91,6 +115,33 @@ When onboarding a new tracing project to Engine, capture the nature of that proj
 
 A server keeps the first workspace's tracing configuration for its lifetime. Workspaces sharing that server must use matching tracing credentials, projects, endpoints, profile selectors, tracing flags, and redaction policy. A conflicting workspace is refused before its settings can affect cached or concurrent runs; start a separate server for that workspace. Restart the server to change its tracing configuration.
 
+## Idle context compaction
+
+Idle compaction is disabled by default. When enabled, the interactive client
+automatically uses the existing `/offload` operation for contexts at or above
+the configured token threshold during the final 20% of a known prompt-cache
+retention window, capped at 120 seconds (60 seconds for a 5-minute window). This keeps the same thread and archives older
+messages; it does not create a handoff thread. The configured offload model is
+used, so summarization can incur costs and cache reuse is not guaranteed.
+
+Opt in with a token threshold through the normal layered config:
+
+```toml
+[history]
+idle_compaction_tokens = 100000
+```
+
+Alternatively, set `DEEPAGENTS_CODE_IDLE_COMPACTION_TOKENS=100000` before
+launching dcode.
+
+Set `idle_compaction_tokens = 0` to disable it again. File edits take effect after
+`/reload`. This setting is independent of cold-cache warning preferences.
+Compaction skips busy sessions, queued input, active typing, modals, lifecycle
+transitions, and approximate token counts. It attempts at most once per thread
+and cache window, announces the operation, and supports Escape cancellation.
+Input arriving before the worker starts skips compaction; input arriving after
+it starts uses the normal offload queue and runs once compaction finishes.
+
 ## Debugging
 
 Deep Agents Code runs as two processes: the **Textual TUI** you interact with, and a **`langgraph dev` subprocess** that hosts the agent graph. Each writes its own log, and a single switch turns both on:

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: deepagents-code
-Version: 0.1.86
+Version: 0.1.88
 Summary: Terminal coding agent built on the Deep Agents SDK - works with any model, with persistent memory, customizable skills, subagents, and approval controls.
 Project-URL: Homepage, https://www.langchain.com/dcode
 Project-URL: Documentation, https://docs.langchain.com/oss/deepagents/code/overview
@@ -30,7 +30,7 @@ Requires-Python: <4.0,>=3.12
 Requires-Dist: aiosqlite<1.0.0,>=0.22.1
 Requires-Dist: anyio<5.0.0,>=4.14.2
 Requires-Dist: deepagents-acp<1.0.0,>=0.0.10
-Requires-Dist: deepagents==0.7.28
+Requires-Dist: deepagents==0.7.30
 Requires-Dist: fastmcp<5.0.0,>=4.0.0b4
 Requires-Dist: filelock<3.30,>=3.29.7
 Requires-Dist: genai-prices<0.2.0,>=0.1.13
@@ -41,14 +41,14 @@ Requires-Dist: langchain-openai<2.0.0,>=1.7.2
 Requires-Dist: langchain-quickjs<0.4.0,>=0.3.4
 Requires-Dist: langchain<2.0.0,>=1.4.6
 Requires-Dist: langgraph-checkpoint-sqlite<4.0.0,>=3.1.2
-Requires-Dist: langgraph-cli[inmem]<1.0.0,>=0.4.35
-Requires-Dist: langgraph-runtime-inmem<1.0.0,>=0.36.1
-Requires-Dist: langgraph-sdk<1.0.0,>=0.4.7
+Requires-Dist: langgraph-cli[inmem]<1.0.0,>=0.4.37
+Requires-Dist: langgraph-runtime-inmem<1.0.0,>=0.36.2
+Requires-Dist: langgraph-sdk<1.0.0,>=0.4.8
 Requires-Dist: langsmith[sandbox]>=0.14.11
 Requires-Dist: markdownify<2.0.0,>=1.2.3
 Requires-Dist: mcp<3.0.0,>=2.0.0
 Requires-Dist: packaging>=26.2
-Requires-Dist: pillow<13.0.0,>=12.3.0
+Requires-Dist: pillow<14.0.0,>=13.0.0
 Requires-Dist: prompt-toolkit<4.0.0,>=3.0.53
 Requires-Dist: pydantic<2.14.0a0,>=2.12.5
 Requires-Dist: pyperclip<2.0.0,>=1.11.0
@@ -120,7 +120,7 @@ Provides-Extra: litellm
 Requires-Dist: langchain-litellm<2.0.0,>=0.11.1; extra == 'litellm'
 Provides-Extra: media
 Requires-Dist: av<19.0.0,>=18.0.0; extra == 'media'
-Requires-Dist: pillow<13.0.0,>=12.3.0; extra == 'media'
+Requires-Dist: pillow<14.0.0,>=13.0.0; extra == 'media'
 Provides-Extra: meta
 Requires-Dist: langchain-meta<2.0.0,>=0.5.0; extra == 'meta'
 Provides-Extra: mistralai

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/deepagents_code/_build_info.py

@@ -1,3 +1,3 @@
 """Generated at build time. Do not edit or commit."""
 
-BUILD_COMMIT = "577e3f1"
+BUILD_COMMIT = "0324ff3"

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/deepagents_code/_constants.py

@@ -14,7 +14,7 @@ DEFAULT_AGENT_NAME: Final[str] = "agent"
 """Default agent / assistant identifier when no `-a` flag is given."""
 
 DEFAULT_THREAD_LIMIT: Final[int] = 100
-"""Fallback maximum number of recent threads to load and display."""
+"""Fallback for `get_thread_limit()`, the thread switcher batch size."""
 
 FS_TOOL_NAMES: Final[frozenset[str]] = frozenset(
     {"ls", "read_file", "write_file", "edit_file", "delete", "glob", "grep", "execute"}

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/deepagents_code/_env_vars.py

@@ -328,6 +328,9 @@ HIDE_SPLASH_TIPS = "DEEPAGENTS_CODE_HIDE_SPLASH_TIPS"
 HIDE_SPLASH_VERSION = "DEEPAGENTS_CODE_HIDE_SPLASH_VERSION"
 """Hide version and local-install details in the splash screen when enabled."""
 
+IDLE_COMPACTION_TOKENS = "DEEPAGENTS_CODE_IDLE_COMPACTION_TOKENS"
+"""Minimum context tokens for opt-in idle compaction; 0 disables it."""
+
 HISTORY_RETENTION_DAYS = "DEEPAGENTS_CODE_HISTORY_RETENTION_DAYS"
 """Days an offloaded conversation-history archive is kept before the startup
 sweep deletes it.
@@ -514,7 +517,11 @@ repo file, so a project `.env` cannot disable itself.
 """
 
 RECENT_THREADS = "DEEPAGENTS_CODE_RECENT_THREADS"
-"""Maximum recent threads (default: `_constants.DEFAULT_THREAD_LIMIT`)."""
+"""Recent threads per batch (default: 100, `_constants.DEFAULT_THREAD_LIMIT`).
+
+The thread switcher loads this many at a time. It is also the default row
+limit for `threads list` (override with `-n`) and `@@` thread completion.
+"""
 
 RECURSION_LIMIT = "DEEPAGENTS_CODE_RECURSION_LIMIT"
 """Override the main agent's LangGraph `recursion_limit` (graph step budget).

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/deepagents_code/_version.py

@@ -2,7 +2,7 @@
 
 # Keep the `x-release-please-version` annotation — release-please uses it to
 # bump `__version__` in sync with `pyproject.toml` on every release PR.
-__version__ = "0.1.86"  # x-release-please-version
+__version__ = "0.1.88"  # x-release-please-version
 
 DOCS_URL = "https://docs.langchain.com/oss/python/deepagents/code"
 """URL for `deepagents-code` documentation."""

{deepagents_code-0.1.86 → deepagents_code-0.1.88}/deepagents_code/agent.py

@@ -1160,6 +1160,7 @@ def load_async_subagents(config_path: Path | None = None) -> list[AsyncSubAgent]
     description = "Research agent"
     url = "https://my-deployment.langsmith.dev"
     graph_id = "agent"
+    webhook = "https://example.com/webhooks/subagent-finished"
     ```
 
     Args:
@@ -1225,6 +1226,8 @@ def load_async_subagents(config_path: Path | None = None) -> list[AsyncSubAgent]
         }
         if "url" in spec and isinstance(spec["url"], str):
             agent["url"] = spec["url"]
+        if "webhook" in spec and isinstance(spec["webhook"], str):
+            agent["webhook"] = spec["webhook"]
         if "headers" in spec and isinstance(spec["headers"], dict):
             agent["headers"] = spec["headers"]
         agents.append(agent)