fix(agents): restrict inline-code sandbox to a json-only __import__ #118

Merged

CoreRasurae merged 1 commit from bugfix/m1-inline-sandbox-import-restriction into master 2026-08-07 15:41:26 +00:00

Conversation 3 Commits 1 Files changed 16 +268 -300

CoreRasurae commented 2026-08-06 17:55:15 +00:00

Member

Copy link

Summary

ToolAgent._execute_python_code bound the inline-code sandbox's __builtins__["__import__"] directly to the real, unrestricted __import__ builtin, so inline code (a type: tool agent's code: body, or the exec_python-gated python_exec tool) could import os/subprocess/socket and reach exactly the filesystem/network/process facilities docs/index.md §13.2.1's exhaustive Restricted Built-ins table and §13.2.3's prohibited-capabilities list were written to keep out.

Fixes it by replacing the raw reference with a new restricted shim, _sandboxed_import, that permits only json and rejects every other module name via the same NameError → ExecutionError translation _execute_python_code already uses for other prohibited-name access.

Collateral fixes required to keep the suite green (both depended on the exact gap being closed):

• Retired features/stderr_suppression.feature/steps: it exercised inline code doing import sys to reach real stderr, which is no longer reachable by design.
• Rewrote regex-based (import re) text extraction in several tests/fixtures/email_graph/*.yaml agent configs to use only sandbox-permitted str methods, preserving the same output keys.

Test plan

• nox -s unit_tests — 2962 scenarios passed, including new positive (import json) and TDD regression (@tdd_issue_107, @tdd_expected_fail removed) scenarios in features/inline_sandbox_import_restriction.feature.
• nox -s integration_tests — 345 Robot tests passed, including new robot/inline_sandbox_import_restriction.robot.
• nox -s coverage_report — COVERAGE OK: 96.6% (threshold 96.5%).
• nox -s lint, nox -s format -- --check, nox -s typecheck, nox -s security_scan, nox -s dead_code — all green.
• Manually verified import json / from json import dumps succeed and import os/subprocess raise ExecutionError via the sandbox.

Closes #107

## Summary `ToolAgent._execute_python_code` bound the inline-code sandbox's `__builtins__["__import__"]` directly to the real, unrestricted `__import__` builtin, so inline code (a `type: tool` agent's `code:` body, or the `exec_python`-gated `python_exec` tool) could `import os`/`subprocess`/`socket` and reach exactly the filesystem/network/process facilities docs/index.md §13.2.1's exhaustive Restricted Built-ins table and §13.2.3's prohibited-capabilities list were written to keep out. Fixes it by replacing the raw reference with a new restricted shim, `_sandboxed_import`, that permits only `json` and rejects every other module name via the same `NameError` → `ExecutionError` translation `_execute_python_code` already uses for other prohibited-name access. **Collateral fixes required to keep the suite green** (both depended on the exact gap being closed): - Retired `features/stderr_suppression.feature`/steps: it exercised inline code doing `import sys` to reach real stderr, which is no longer reachable by design. - Rewrote regex-based (`import re`) text extraction in several `tests/fixtures/email_graph/*.yaml` agent configs to use only sandbox-permitted `str` methods, preserving the same output keys. ## Test plan - [x] `nox -s unit_tests` — 2962 scenarios passed, including new positive (`import json`) and TDD regression (`@tdd_issue_107`, `@tdd_expected_fail` removed) scenarios in `features/inline_sandbox_import_restriction.feature`. - [x] `nox -s integration_tests` — 345 Robot tests passed, including new `robot/inline_sandbox_import_restriction.robot`. - [x] `nox -s coverage_report` — COVERAGE OK: 96.6% (threshold 96.5%). - [x] `nox -s lint`, `nox -s format -- --check`, `nox -s typecheck`, `nox -s security_scan`, `nox -s dead_code` — all green. - [x] Manually verified `import json` / `from json import dumps` succeed and `import os`/`subprocess` raise `ExecutionError` via the sandbox. Closes #107

CoreRasurae added this to the v2.1.0 milestone 2026-08-06 17:55:15 +00:00

CoreRasurae added the

Type

Bug

label 2026-08-06 18:01:53 +00:00

CoreRasurae added a new dependency 2026-08-06 18:02:28 +00:00

#107 Inline-code sandbox exposes __import__, letting code bypass the Restricted Built-ins set entirely (e.g. import os)

hurui200320 approved these changes 2026-08-07 04:48:03 +00:00

hurui200320 left a comment

Copy link

PR Review: !118 (Ticket #107)

Verdict: Approve

The change correctly closes the security gap described in #107: the inline-code sandbox now routes every import statement through _sandboxed_import, which rejects any module other than json and surfaces the failure through the existing NameError → ExecutionError path. Both entry points (type: tool inline code: bodies and the exec_python-gated python_exec built-in tool) are covered, the expression sandbox is untouched, existing import json fixtures continue to work, and the collateral fixture updates remove the now-prohibited import re usage without introducing new imports.

Critical Issues

None

Major Issues

None

Minor Issues

• Hardening: _sandboxed_import does not reject non-zero level arguments.
  File: src/cleveractors/agents/tool.py, lines 35–53
  The shim correctly blocks every module name except json, but it still passes a non-zero level through to the real __import__. For defense-in-depth, consider rejecting level != 0 with the same NameError, since the spec only intends the top-level json module to be reachable and relative imports are not part of the allowed surface.

• The regression assertion only checks the error category, not the failure cause.
  File: features/steps/inline_sandbox_import_restriction_steps.py, lines 21–35
  The step asserts isinstance(context.error, ExecutionError), which is good, but it does not verify that the ExecutionError was actually triggered by the forbidden import (e.g., by checking that the module name appears in the message or that the post-import statement did not run). Consider tightening the assertion so a future regression that accidentally allows the import but fails later cannot masquerade as a passing test.

Nits

• Duplicated string-parsing logic in email fixtures.
  Files: tests/fixtures/email_graph/*/{business_email,component_orders}.yaml
  The regex replacement logic for component-order part numbers and quantities is copy-pasted across six fixture files. If these fixtures need further semantic tweaks, a shared helper or template would reduce drift.

• Robot library couples to a private method.
  File: robot/InlineSandboxImportRestrictionLib.py, lines 46–48
  Run Inline Code calls agent._python_exec_tool directly. This is fine for an integration test, but it creates a brittle coupling to a private API; consider documenting it or using the public process_message path.

Summary

This is a focused, well-scoped security fix. The shim is minimal, the regression coverage exercises both documented entry points and multiple dynamic-import forms, and the collateral cleanup (removing import re from fixtures and retiring the stderr_suppression feature that depended on import sys) is clearly justified. No critical or major issues were found; the minor items above are optional hardening/quality improvements.

## PR Review: !118 (Ticket #107) ### Verdict: Approve The change correctly closes the security gap described in #107: the inline-code sandbox now routes every `import` statement through `_sandboxed_import`, which rejects any module other than `json` and surfaces the failure through the existing `NameError` → `ExecutionError` path. Both entry points (`type: tool` inline `code:` bodies and the `exec_python`-gated `python_exec` built-in tool) are covered, the expression sandbox is untouched, existing `import json` fixtures continue to work, and the collateral fixture updates remove the now-prohibited `import re` usage without introducing new imports. ### Critical Issues None ### Major Issues None ### Minor Issues - **Hardening: `_sandboxed_import` does not reject non-zero `level` arguments.** File: `src/cleveractors/agents/tool.py`, lines 35–53 The shim correctly blocks every module name except `json`, but it still passes a non-zero `level` through to the real `__import__`. For defense-in-depth, consider rejecting `level != 0` with the same `NameError`, since the spec only intends the top-level `json` module to be reachable and relative imports are not part of the allowed surface. - **The regression assertion only checks the error category, not the failure cause.** File: `features/steps/inline_sandbox_import_restriction_steps.py`, lines 21–35 The step asserts `isinstance(context.error, ExecutionError)`, which is good, but it does not verify that the `ExecutionError` was actually triggered by the forbidden import (e.g., by checking that the module name appears in the message or that the post-import statement did not run). Consider tightening the assertion so a future regression that accidentally allows the import but fails later cannot masquerade as a passing test. ### Nits - **Duplicated string-parsing logic in email fixtures.** Files: `tests/fixtures/email_graph/*/{business_email,component_orders}.yaml` The regex replacement logic for component-order part numbers and quantities is copy-pasted across six fixture files. If these fixtures need further semantic tweaks, a shared helper or template would reduce drift. - **Robot library couples to a private method.** File: `robot/InlineSandboxImportRestrictionLib.py`, lines 46–48 `Run Inline Code` calls `agent._python_exec_tool` directly. This is fine for an integration test, but it creates a brittle coupling to a private API; consider documenting it or using the public `process_message` path. ### Summary This is a focused, well-scoped security fix. The shim is minimal, the regression coverage exercises both documented entry points and multiple dynamic-import forms, and the collateral cleanup (removing `import re` from fixtures and retiring the `stderr_suppression` feature that depended on `import sys`) is clearly justified. No critical or major issues were found; the minor items above are optional hardening/quality improvements.

CoreRasurae force-pushed bugfix/m1-inline-sandbox-import-restriction from 64cbb46f35 to e99eb26579 2026-08-07 14:08:58 +00:00 Compare

CoreRasurae referenced this pull request from a commit 2026-08-07 14:08:58 +00:00

fix(agents): restrict inline-code sandbox to a json-only __import__

CoreRasurae commented 2026-08-07 14:10:01 +00:00

Author

Member

Copy link

Thanks for the review, @hurui200320! Addressed the minor issues; the two nits were left as-is with reasoning below.

Applied:

• _sandboxed_import now also rejects a non-zero level (relative import forms such as from .json import dumps) — §13.2.1 lists only the absolute json module, so relative forms aren't part of the allowed surface. Added a new scenario (python_exec tool rejects a relative import of json) covering this path.
• The regression assertion in inline_sandbox_import_restriction_steps.py now also checks that the rejected module name appears in the ExecutionError message and that result stays unset (the statement after the rejected import never ran), so it can no longer pass on an unrelated error.

Left as-is (nits, not spec violations):

• Duplicated str-based extraction logic across the email_graph fixture YAMLs: genuine deduplication isn't feasible here without either introducing a shared import the sandbox is specifically designed to forbid, or an unrelated fixture-templating refactor — disproportionate scope for this PR. Happy to open a follow-up issue if you'd like it tracked separately.
• robot/InlineSandboxImportRestrictionLib.py calling agent._python_exec_tool directly: test-only code, not a maintainability concern for the production API surface.

All nox sessions green (lint, format, typecheck, security_scan, dead_code, unit_tests, coverage delta on the touched code, integration_tests, e2e_tests). Amended the existing commit (still unmerged, single-commit branch) rather than adding a new one, per project convention.

Thanks for the review, @hurui200320! Addressed the minor issues; the two nits were left as-is with reasoning below. **Applied:** - `_sandboxed_import` now also rejects a non-zero `level` (relative import forms such as `from .json import dumps`) — §13.2.1 lists only the absolute `json` module, so relative forms aren't part of the allowed surface. Added a new scenario (`python_exec tool rejects a relative import of json`) covering this path. - The regression assertion in `inline_sandbox_import_restriction_steps.py` now also checks that the rejected module name appears in the `ExecutionError` message and that `result` stays unset (the statement after the rejected import never ran), so it can no longer pass on an unrelated error. **Left as-is (nits, not spec violations):** - Duplicated str-based extraction logic across the `email_graph` fixture YAMLs: genuine deduplication isn't feasible here without either introducing a shared import the sandbox is specifically designed to forbid, or an unrelated fixture-templating refactor — disproportionate scope for this PR. Happy to open a follow-up issue if you'd like it tracked separately. - `robot/InlineSandboxImportRestrictionLib.py` calling `agent._python_exec_tool` directly: test-only code, not a maintainability concern for the production API surface. All nox sessions green (lint, format, typecheck, security_scan, dead_code, unit_tests, coverage delta on the touched code, integration_tests, e2e_tests). Amended the existing commit (still unmerged, single-commit branch) rather than adding a new one, per project convention.

CoreRasurae requested review from hurui200320 2026-08-07 14:10:21 +00:00

CoreRasurae added 1 commit 2026-08-07 14:30:55 +00:00

fix(agents): restrict inline-code sandbox to a json-only __import__ ... c5154701aa

ToolAgent._execute_python_code built the inline-code sandbox's
__builtins__ dict per docs/index.md §13.2.1's "Restricted Built-ins for
Inline Code" table (exhaustive per that section, and per §13.2.3's
"dynamic import of modules other than those explicitly listed" is
prohibited), but bound "__import__" directly to the real, unrestricted
__import__ builtin. Since __import__ is the mechanism the `import`
statement invokes, inline code -- a "type: tool" agent's `code:` body,
or the exec_python-gated python_exec tool, both of which share this
sandbox construction -- could write `import os` (or any other module)
and reach exactly the filesystem/network/process facilities the
standard was written to keep out.

Fixes it by replacing the raw __import__ reference with a new
module-level shim, _sandboxed_import, that permits only `json` (the
sole module-shaped facility §13.2.1 lists) and rejects every other
module name via NameError -- the same category _execute_python_code
already uses for other prohibited-name access, translated to
ExecutionError, rather than a bare Python traceback. `import json`
keeps working unchanged; the §13.2.2 expression sandbox is unaffected,
since it never exposed __import__.

Two collateral fixes were required to keep the existing suite green,
both because they depended on the exact `__import__` gap being closed:

- features/stderr_suppression.feature exercised inline code doing
  `import sys; print(..., file=sys.stderr)` to validate a debug-output
  suppression path in _execute_python_code. `sys` is not in §13.2.1's
  table, so this is now unreachable by design -- there is no longer any
  way for inline code to obtain a real stderr handle. Retired the
  feature and its steps; the underlying redirect logic in tool.py is
  untouched (harmless, and still guards defense-in-depth).

- Several tests/fixtures/email_graph/*.yaml agent configs used
  `import json, re` for regex-based text extraction (part numbers,
  quantities, meeting mentions). `re` is likewise not in §13.2.1's
  table. Rewrote the extraction logic in each affected fixture using
  only sandbox-permitted str methods (no regex needed for these
  fixed-shape patterns), preserving the same output keys.

Regression scenarios (tagged @tdd_issue, @tdd_issue_107) proving this
bug were merged separately in #108; this commit removes the
@tdd_expected_fail tag now that the fix is in place, and adds a
positive scenario plus a Robot integration test confirming `import
json` still works while os/socket/subprocess imports are rejected
end-to-end.

Addressed review feedback from @hurui200320 on PR !118:

- _sandboxed_import now also rejects a non-zero `level` (relative
  import forms such as `from .json import dumps`), since §13.2.1 lists
  only the absolute `json` module and relative imports are not part of
  that allowed surface. A new scenario in
  inline_sandbox_import_restriction.feature covers this path.
- The regression assertion (inline_sandbox_import_restriction_steps.py)
  now also checks that the rejected module name appears in the
  ExecutionError message and that the statement following the rejected
  import never ran (`result` stays unset), so the assertion can no
  longer pass on an unrelated error.
- Deduplicated the str-based extraction logic that was copy-pasted
  across seven tests/fixtures/email_graph/*.yaml agent configs (both
  the meeting/urgency detection and the part/quantity/BOM detection).
  Since the sandboxed `code:` bodies still cannot import a shared
  helper module (that is the exact restriction this commit adds), the
  two snippets now live once each as constants in robot/EmailGraphLib.py
  and are expanded from a placeholder token back into every fixture's
  parsed config right after load (yaml.safe_load, and after resolving a
  namespaced package via LocalPackageStore), so every consumer still
  executes byte-identical Python source assembled from one canonical
  copy. Verified equivalence by executing the old and new snippets
  against representative inputs and diffing the results, and re-ran the
  Robot email_graph suite end-to-end (345/345 passed).
- Left as-is (reviewer nit, not a spec violation): the Robot library's
  use of ToolAgent's private _python_exec_tool -- test-only code, not a
  maintainability concern for the production API surface.

ISSUES CLOSED: #107

CoreRasurae force-pushed bugfix/m1-inline-sandbox-import-restriction from e99eb26579 to c5154701aa 2026-08-07 14:30:55 +00:00 Compare

CoreRasurae referenced this pull request from a commit 2026-08-07 14:30:55 +00:00

fix(agents): restrict inline-code sandbox to a json-only __import__

CoreRasurae commented 2026-08-07 15:35:03 +00:00

Author

Member

Copy link

Follow-up: went ahead and deduplicated the fixture extraction logic too (code quality/maintainability took priority over leaving it as a nit).

Since the sandboxed code: bodies still can't import a shared helper module (that's the exact restriction this PR adds), the two duplicated snippets (business_email meeting/urgency detection, component_orders part/quantity/BOM detection) now live once each as constants in robot/EmailGraphLib.py, expanded from a placeholder token back into each fixture's config right after yaml.safe_load (and after resolving a namespaced package via LocalPackageStore) — before the config ever reaches create_executor. Verified equivalence by executing the old and new snippets against representative inputs and diffing the results, then re-ran the full Robot email_graph suite end-to-end (345/345 passed).

The private-method-coupling nit (InlineSandboxImportRestrictionLib.py calling agent._python_exec_tool) is still left as-is — test-only code, not a maintainability concern for the production API.

All nox sessions re-verified green after this change. Force-pushed the amended commit again.

Follow-up: went ahead and deduplicated the fixture extraction logic too (code quality/maintainability took priority over leaving it as a nit). Since the sandboxed `code:` bodies still can't import a shared helper module (that's the exact restriction this PR adds), the two duplicated snippets (business_email meeting/urgency detection, component_orders part/quantity/BOM detection) now live once each as constants in `robot/EmailGraphLib.py`, expanded from a placeholder token back into each fixture's config right after `yaml.safe_load` (and after resolving a namespaced package via `LocalPackageStore`) — before the config ever reaches `create_executor`. Verified equivalence by executing the old and new snippets against representative inputs and diffing the results, then re-ran the full Robot email_graph suite end-to-end (345/345 passed). The private-method-coupling nit (`InlineSandboxImportRestrictionLib.py` calling `agent._python_exec_tool`) is still left as-is — test-only code, not a maintainability concern for the production API. All nox sessions re-verified green after this change. Force-pushed the amended commit again.

CoreRasurae merged commit c5154701aa into master 2026-08-07 15:41:26 +00:00

CoreRasurae deleted branch bugfix/m1-inline-sandbox-import-restriction 2026-08-07 15:41:34 +00:00

hurui200320 referenced this pull request 2026-08-08 07:06:45 +00:00

test(agents): capture LLM-agent file_write unsafe-mode regression (#115) #119

Sign in to join this conversation.

Reviewers

No reviewers

hurui200320

Labels

Clear labels

  

auto/blocked-by-deps

PR blocked by an open issue dependency. Operator must close the dep (or remove the dependency link) before the merge driver can act. Auto-cleared by merge_drive when no open deps remain.

  

auto/ci-timeout

Most recent merge cycle hit CI timeout. Driver excludes this PR while last merge_cycle row is < 30 min old; label persists thereafter as visible history.

  

auto/claimed-implementer

Currently being processed by an implementer worker.

  

auto/claimed-merge

Currently being processed by the merge driver.

  

auto/claimed-reviewer

Currently being processed by a reviewer worker.

  

auto/driver-down

Merge driver heartbeat stale; pipeline halted. Closed automatically on next clean tick.

  

auto/invariant-violation

Detected master commit violating the strict merge invariant. Tracked as an issue (not a PR label); kept here for label completeness.

  

auto/last-attempt-tier-0

In-cycle escalation: most recent attempt ran at the Tier 0 slot (`tier-0`). Slot's model defined in .opencode/models/tiers.yaml.

  

auto/last-attempt-tier-1

In-cycle escalation: most recent attempt ran at the Tier 1 slot (`tier-1`). Slot's model defined in .opencode/models/tiers.yaml.

  

auto/last-attempt-tier-2

In-cycle escalation: most recent attempt ran at the Tier 2 slot (`tier-2`). Slot's model defined in .opencode/models/tiers.yaml. Gated behind IMPLEMENTER_ESCALATION_TIER2_ENABLED.

  

auto/last-attempt-tier-min

In-cycle escalation: most recent attempt ran at the Tier -1 slot (`tier-min`). Slot's model defined in .opencode/models/tiers.yaml. Suffix is ``-min`` (not ``--1``) so the Forgejo UI reads naturally.

  

Automation Tracking

Tracking issues used by the AI Automation system for agents to communicate and report.

  

auto/needs-conflict-resolution

Rebase conflict needs LLM conflict-resolver.

  

auto/needs-implementer

Failing CI needs implementer attention.

  

auto/postmortem

Documenting a driver incident or rollback.

  

auto/ready-to-merge

Reviewer has APPROVED this PR and no later REQUEST_CHANGES is outstanding. The merge driver requires this label to even consider a PR for merging. Set by the reviewer worker on APPROVE; cleared on REQUEST_CHANGES.

  

auto/restart-throttled

Train repeatedly lost master-tempo races. Driver excludes via merge_cycle until cooldown elapses; label persists as visible history.

  

auto/revert

Revert PR backing out an invariant violation. Fast-tracked through the merge driver.

  

auto/sentinel

Sentinel PR duplicated from upstream into a personal fork by tools/duplicate_prs_to_fork.py for pipeline testing. Lives only in the fork; the canonical pipeline never sees it.

  

auto/stale-inactivity

No implementer activity for N days. Flagged for human review. Auto-cleared on next push to head branch.

  

auto/unstable

Repeatedly fails on current master (>= 3 ci-fail-on-rebased-sha releases in 12 h). Excluded from driver until human triage.

  

Blocked

A ticket in a blocked state and unable to complete until some other task is completed first.

  

Bounty

$100

A bounty of $100 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$1000

A bounty of $1000 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$10000

A bounty of $10000 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$20

A bounty of $20 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$2000

A bounty of $2000 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$250

A bounty of $250 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$50

A bounty of $50 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$500

A bounty of $500 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$5000

A bounty of $5000 for any open-source contributor who provides a MR that solves this issue

  

Bounty

$750

A bounty of $750 for any open-source contributor who provides a MR that solves this issue

  

MoSCoW

Could have

Could have feature in order to satisfy the epic/legendary.

  

MoSCoW

Must have

Must have feature in order to satisfy the epic/legendary.

  

MoSCoW

Should have

Should have feature in order to satisfy the epic/legendary.

  

Needs Feedback

There are questions in the ticket that can not be completed until the project owner provides clarity.

  

Points

1

1 man-hours worth of work for an expert with no learning curve.

  

Points

13

13 man-hours worth of work for an expert with no learning curve.

  

Points

2

2 man-hours worth of work for an expert with no learning curve.

  

Points

21

21 man-hours worth of work for an expert with no learning curve.

  

Points

3

3 man-hours worth of work for an expert with no learning curve.

  

Points

34

34 man-hours worth of work for an expert with no learning curve.

  

Points

5

5 man-hours worth of work for an expert with no learning curve.

  

Points

55

55 man-hours worth of work for an expert with no learning curve.

  

Points

8

8 man-hours worth of work for an expert with no learning curve.

  

Points

88

88 man-hours worth of work for an expert with no learning curve.

  

Priority

Backlog

This ticket has backlogged priority and is not to be worked on yet

  

Priority

CI Blocker

Critical priority issue that blocks CI/CD pipeline and prevents PR merges

  

Priority

Critical

The priority is critical

  

Priority

High

The priority is high

  

Priority

Low

The priority is low

  

Priority

Medium

The priority is medium

  

Signed-off: Owner

When an epic or legendary is in review it must be signed off by owner, tech lead, and scrum master before being marked as completed.

  

Signed-off: Scrum Master

When an epic or legendary is in review it must be signed off by owner, tech lead, and scrum master before being marked as completed.

  

Signed-off: Tech Lead

When an epic or legendary is in review it must be signed off by owner, tech lead, and scrum master before being marked as completed.

  

Spike

A ticket for learning a tool or technology that is needed to be able to do future planning and design.

  

State

Completed

The ticket has been fully implemented, completed, and merged with the source code. This label should only be applied once a ticket is closed.

  

State

Duplicate

A ticket that represents the same content as an existing ticket.

  

State

In Progress

A ticket that is actively being developed.

  

State

In Review

A ticket that has had some code completed to implement but is waiting to pass peer review and is not yet merged in.

  

State

Paused

This ticket's work started but wasn't finished. It's on hold (likely in a feature branch) and will be resumed later, either due to a blocker or a delay.

  

State

Unverified

All new tickets start in this state. A developer may set it to show the ticket is unverified. This means we haven't agreed to work on it. It will either move to a verified state or be closed as wontdo.

  

State

Verified

The issue has been verified by a developer as legitimate. It will be worked on and verified tickets are now considered part of the backlog.

  

State

Wont Do

This ticket has been decided it wont be done. This may mean the bug has been determined to not be real (cant verify) or the feature is one we have decided we dont want to adopt.

  

Type

Automation

Any edits or discussion about the AI automated coding system.

  

Type

Bug

Something that doesnt work as intended.

  

Type

Discussion

Anytime a ticket represents a discussion about a subject and doesnt fall into one of the other categories.

  

Type

Documentation

An error or improvement needed in the documentation.

  

Type

Epic

Any first tier epic. That is, an epic which contains only issues as children and will not have sub-epics.

  

Type

Feature

Some new functionality not present.

  

Type

Legendary

A type of Epic which will contain other Epics.

  

Type

Refactor

A code change that restructures existing code without changing its external behavior.

  

Type

Support

Someone needs help using the project.

  

Type

Task

A generic task that doesnt fit into the other type categories.

  

Type

Testing

Work exclusively focusing on fixing or expanding testing.

No labels

auto/blocked-by-deps

auto/ci-timeout

auto/claimed-implementer

auto/claimed-merge

auto/claimed-reviewer

auto/driver-down

auto/invariant-violation

auto/last-attempt-tier-0

auto/last-attempt-tier-1

auto/last-attempt-tier-2

auto/last-attempt-tier-min

Automation Tracking

auto/needs-conflict-resolution

auto/needs-implementer

auto/postmortem

auto/ready-to-merge

auto/restart-throttled

auto/revert

auto/sentinel

auto/stale-inactivity

auto/unstable

Blocked

Bounty

$100

Bounty

$1000

Bounty

$10000

Bounty

$20

Bounty

$2000

Bounty

$250

Bounty

$50

Bounty

$500

Bounty

$5000

Bounty

$750

MoSCoW

Could have

MoSCoW

Must have

MoSCoW

Should have

Needs Feedback

Points

1

Points

13

Points

2

Points

21

Points

3

Points

34

Points

5

Points

55

Points

8

Points

88

Priority

Backlog

Priority

CI Blocker

Priority

Critical

Priority

High

Priority

Low

Priority

Medium

Signed-off: Owner

Signed-off: Scrum Master

Signed-off: Tech Lead

Spike

State

Completed

State

Duplicate

State

In Progress

State

In Review

State

Paused

State

Unverified

State

Verified

State

Wont Do

Type

Automation

Type

Bug

Type

Discussion

Type

Documentation

Type

Epic

Type

Feature

Type

Legendary

Type

Refactor

Type

Support

Type

Task

Type

Testing

Milestone

Clear milestone

No items

No milestone

v2.1.0

Projects

Clear projects

No items

No project

Assignees

Clear assignees

No assignees

2 participants

Notifications

Subscribe

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Blocks

#107 Inline-code sandbox exposes __import__, letting code bypass the Restricted Built-ins set entirely (e.g. import os)

cleveragents/cleveractors-core

Reference

cleveragents/cleveractors-core!118

No description provided.