Skip to content

fix(code): put local offloaded tool results on the real filesystem - #4740

Merged
Mason Daugherty (mdrxy) merged 7 commits into
mainfrom
mdrxy/code/local-tmp-artifacts-root
Jul 15, 2026
Merged

fix(code): put local offloaded tool results on the real filesystem#4740
Mason Daugherty (mdrxy) merged 7 commits into
mainfrom
mdrxy/code/local-tmp-artifacts-root

Conversation

@mdrxy

@mdrxy Mason Daugherty (mdrxy) commented Jul 14, 2026 •
edited
Loading

Copy link
Copy Markdown
Member

In dcode local mode, offloaded large tool results previously lived behind a virtual CompositeBackend route backed by a hidden mkdtemp FilesystemBackend. Because execute always runs on the default backend (never path-routed), the agent could not open those results with jq/grep/python by the path it was handed — the /large_tool_results/ prefix wasn't a real inode, and the temp root was intentionally hidden.

This points CompositeBackend.artifacts_root at a stable, hardened per-user temp dir and drops the /large_tool_results/ route, so offloaded results fall through to the default LocalShellBackend at a real path. The agent can now inspect them with execute using the exact path the offload message hands it — no virtual-path translation, no copy-out/in. Conversation history keeps a dedicated route to persistent ~/.deepagents storage so /offload archives survive restarts, and its address stays stable across restarts because the artifacts root is deterministic per user.

This uses the existing artifacts_root support on CompositeBackend (already consumed by both the filesystem and summarization middleware), so it is a dcode-only change with no SDK edit. The artifacts dir is created 0o700 with an ownership/S_ISDIR guard and falls back to a private unique dir if the predictable path is squatted; consolidating on one per-user dir also removes the old unbounded mkdtemp-per-session leak.

The rubric grader's read allow-list now derives its permitted prefix from artifacts_root rather than a hardcoded /large_tool_results/.

Made by Open SWE

References

@mdrxy @open-swe
feat(code): put local offloaded tool results on the real filesystem
6fc8be0 
In local mode, point `CompositeBackend.artifacts_root` at a stable, hardened
per-user temp dir and drop the virtual `/large_tool_results/` route so offloaded
tool results fall through to the default `LocalShellBackend` at a real path. The
agent can now inspect them with `execute` (jq/grep/python) using the exact path
the offload message hands it -- no virtual-path translation or copy-out/in.
Conversation history keeps a dedicated route to persistent `~/.deepagents`
storage. Uses the existing `CompositeBackend.artifacts_root` support, so this is
a dcode-only change with no SDK edit.

Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
@github-actions github-actions Bot added dcode Related to `deepagents-code` feature New feature/enhancement or request for one internal User is a member of the `langchain-ai` GitHub organization size: M 200-499 LOC labels Jul 14, 2026
@mdrxy
Mason Daugherty (mdrxy) marked this pull request as ready for review July 14, 2026 22:11
@mdrxy Mason Daugherty (mdrxy) changed the title feat(code): put local offloaded tool results on the real filesystem fix(code): put local offloaded tool results on the real filesystem Jul 14, 2026
@github-actions github-actions Bot added fix A bug fix (PATCH) and removed feature New feature/enhancement or request for one labels Jul 14, 2026
open-swe[bot]
open-swe Bot reviewed Jul 14, 2026
View reviewed changes

@open-swe open-swe Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Open SWE Review found 2 potential issues.

Open in Web • View Open SWE trace

Comment thread libs/code/deepagents_code/offload.py
Comment thread libs/code/deepagents_code/offload.py Outdated
@github-actions github-actions Bot added size: L 500-999 LOC and removed size: M 200-499 LOC labels Jul 14, 2026
@mdrxy
Mason Daugherty (mdrxy) merged commit 36051e2 into main Jul 15, 2026
52 checks passed
@mdrxy
Mason Daugherty (mdrxy) deleted the mdrxy/code/local-tmp-artifacts-root branch July 15, 2026 13:24
@github-actions github-actions Bot mentioned this pull request Jul 15, 2026
Merged
Mason Daugherty (mdrxy) pushed a commit that referenced this pull request Jul 16, 2026
@github-actions @langchain-oss-automated-triage
release(deepagents-code): 0.1.40 (#4734)
019489e 
> [!CAUTION]
> Merging this PR will automatically publish to **PyPI** and create a
**GitHub release**.

For the full release process, see
[`.github/RELEASING.md`](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md).

---

_Release notes preview: keep this section in sync with the package
`CHANGELOG.md`. The published GitHub release body is extracted from the
merged `CHANGELOG.md` by `release.yml`, not from this PR description._

---


##
[0.1.40](deepagents-code==0.1.39...deepagents-code==0.1.40)
(2026-07-16)

### Features

- Added plugin marketplace support
([#4554](#4554)).
- Added an “always allow” option to the project MCP approval prompt
([#4562](#4562)).
- Improved `/goal` workflows: criteria generation now runs server-side,
YOLO mode auto-accepts criteria, goals complete after satisfied grading,
and goal review editing now supports `Ctrl+X`
([#4754](#4754),
[#4784](#4784),
[#4781](#4781),
[#4780](#4780)).
- Reasoning effort now persists across restarts
([#4728](#4728)).
- Added a toast prompting you to re-paste when a chat paste collapses
([#4742](#4742)).

### Bug Fixes

- Tool calls awaiting approval are now surfaced correctly
([#4739](#4739)).
- Fixed transcript tail hydration when scrolled to the bottom edge
([#4733](#4733)).
- Kept chat input responsive during MCP viewer `Ctrl+R` reconnects
([#4753](#4753)).
- Improved inline free-text prompts by sharing paste handling and
matching primary-input `Ctrl+D` behavior
([#4736](#4736),
[#4729](#4729)).
- Fixed local offloaded tool results to use the real filesystem
([#4740](#4740)).
- Cleaned offloaded history when deleting a thread
([#4751](#4751)).
- Removed duplicated content from the system prompt by overwriting the
base prompt
([#4516](#4516)).
- Closed subprocess transport during install teardown
([#4735](#4735)).
- Added targeted `uv` constraints for prerelease dependencies
([#4744](#4744)).

_End release notes preview._

---

> [!NOTE]
> A **New Contributors** section is appended to the GitHub release notes
automatically at publish time (see [Release
Pipeline](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md#release-pipeline),
step 2).

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: langchain-oss-automated-triage[bot] <248757908+langchain-oss-automated-triage[bot]@users.noreply.github.com>
Marcelo5444 pushed a commit to Marcelo5444/deepagents that referenced this pull request Jul 30, 2026
@mdrxy @open-swe
fix(code): put local offloaded tool results on the real filesystem (l…
ebc8132 
…angchain-ai#4740)

In dcode local mode, offloaded large tool results previously lived
behind a virtual `CompositeBackend` route backed by a hidden `mkdtemp`
`FilesystemBackend`. Because `execute` always runs on the default
backend (never path-routed), the agent could not open those results with
`jq`/`grep`/`python` by the path it was handed — the
`/large_tool_results/` prefix wasn't a real inode, and the temp root was
intentionally hidden.

This points `CompositeBackend.artifacts_root` at a stable, hardened
per-user temp dir and drops the `/large_tool_results/` route, so
offloaded results fall through to the default `LocalShellBackend` at a
real path. The agent can now inspect them with `execute` using the exact
path the offload message hands it — no virtual-path translation, no
copy-out/in. Conversation history keeps a dedicated route to persistent
`~/.deepagents` storage so `/offload` archives survive restarts, and its
address stays stable across restarts because the artifacts root is
deterministic per user.

This uses the existing `artifacts_root` support on `CompositeBackend`
(already consumed by both the filesystem and summarization middleware),
so it is a dcode-only change with no SDK edit. The artifacts dir is
created `0o700` with an ownership/`S_ISDIR` guard and falls back to a
private unique dir if the predictable path is squatted; consolidating on
one per-user dir also removes the old unbounded `mkdtemp`-per-session
leak.

The rubric grader's read allow-list now derives its permitted prefix
from `artifacts_root` rather than a hardcoded `/large_tool_results/`.

Made by [Open
SWE](https://openswe.vercel.app/agents/e815e778-3188-4f2e-1068-e52819c024a7)

## References
- Plan:
https://openswe.vercel.app/agents/e815e778-3188-4f2e-1068-e52819c024a7/plan

---------

Co-authored-by: open-swe[bot] <open-swe@users.noreply.github.com>
Marcelo5444 pushed a commit to Marcelo5444/deepagents that referenced this pull request Jul 30, 2026
@github-actions @langchain-oss-automated-triage
release(deepagents-code): 0.1.40 (langchain-ai#4734)
5b5f28d 
> [!CAUTION]
> Merging this PR will automatically publish to **PyPI** and create a
**GitHub release**.

For the full release process, see
[`.github/RELEASING.md`](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md).

---

_Release notes preview: keep this section in sync with the package
`CHANGELOG.md`. The published GitHub release body is extracted from the
merged `CHANGELOG.md` by `release.yml`, not from this PR description._

---


##
[0.1.40](langchain-ai/deepagents@deepagents-code==0.1.39...deepagents-code==0.1.40)
(2026-07-16)

### Features

- Added plugin marketplace support
([langchain-ai#4554](langchain-ai#4554)).
- Added an “always allow” option to the project MCP approval prompt
([langchain-ai#4562](langchain-ai#4562)).
- Improved `/goal` workflows: criteria generation now runs server-side,
YOLO mode auto-accepts criteria, goals complete after satisfied grading,
and goal review editing now supports `Ctrl+X`
([langchain-ai#4754](langchain-ai#4754),
[langchain-ai#4784](langchain-ai#4784),
[langchain-ai#4781](langchain-ai#4781),
[langchain-ai#4780](langchain-ai#4780)).
- Reasoning effort now persists across restarts
([langchain-ai#4728](langchain-ai#4728)).
- Added a toast prompting you to re-paste when a chat paste collapses
([langchain-ai#4742](langchain-ai#4742)).

### Bug Fixes

- Tool calls awaiting approval are now surfaced correctly
([langchain-ai#4739](langchain-ai#4739)).
- Fixed transcript tail hydration when scrolled to the bottom edge
([langchain-ai#4733](langchain-ai#4733)).
- Kept chat input responsive during MCP viewer `Ctrl+R` reconnects
([langchain-ai#4753](langchain-ai#4753)).
- Improved inline free-text prompts by sharing paste handling and
matching primary-input `Ctrl+D` behavior
([langchain-ai#4736](langchain-ai#4736),
[langchain-ai#4729](langchain-ai#4729)).
- Fixed local offloaded tool results to use the real filesystem
([langchain-ai#4740](langchain-ai#4740)).
- Cleaned offloaded history when deleting a thread
([langchain-ai#4751](langchain-ai#4751)).
- Removed duplicated content from the system prompt by overwriting the
base prompt
([langchain-ai#4516](langchain-ai#4516)).
- Closed subprocess transport during install teardown
([langchain-ai#4735](langchain-ai#4735)).
- Added targeted `uv` constraints for prerelease dependencies
([langchain-ai#4744](langchain-ai#4744)).

_End release notes preview._

---

> [!NOTE]
> A **New Contributors** section is appended to the GitHub release notes
automatically at publish time (see [Release
Pipeline](https://github.com/langchain-ai/deepagents/blob/main/.github/RELEASING.md#release-pipeline),
step 2).

---------

Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
Co-authored-by: langchain-oss-automated-triage[bot] <248757908+langchain-oss-automated-triage[bot]@users.noreply.github.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@open-swe open-swe[bot] open-swe[bot] left review comments

Assignees

No one assigned

Labels

dcode Related to `deepagents-code` fix A bug fix (PATCH) internal User is a member of the `langchain-ai` GitHub organization size: L 500-999 LOC

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

1 participant

@mdrxy