mini-SWE-agent changelog

What's Changed

• fix(agents): count billed calls that fail to parse against cost_limit by @EnesYilmazcode in https://github.com/SWE-agent/mini-swe-agent/pull/915

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.5...v2.4.6

What's Changed

• fix(config): read config files as UTF-8 regardless of locale by @EnesYilmazcode in https://github.com/SWE-agent/mini-swe-agent/pull/892
• fix(models): don't mask malformed tool-call errors as truncation by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/896

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.4...v2.4.5

New contributors

Thanks to simple-zy for raising issue https://github.com/SWE-agent/mini-swe-agent/issues/894

What's Changed

• fix: remove startup-banner emoji that crashes import on non-UTF stdout by @EnesYilmazcode in https://github.com/SWE-agent/mini-swe-agent/pull/886
• fix(benchmarks): don't mutate shared config across worker threads in get_sb_environment by @EnesYilmazcode in https://github.com/SWE-agent/mini-swe-agent/pull/887

New Contributors

• @EnesYilmazcode made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/886

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.3...v2.4.4

What's Changed

Fixes

• fix(run): pass the startup_command as a dict to env.execute() function. by @Ethan-Kuang-6666 in https://github.com/SWE-agent/mini-swe-agent/pull/880

Enhancements

• Reject empty config override keys by @CodingFeng101 in https://github.com/SWE-agent/mini-swe-agent/pull/883

New Contributors

• @Ethan-Kuang-6666 made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/880
• @CodingFeng101 made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/883

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.2...v2.4.3

What's Changed

Small release to bump image versions in programbench runner.

• Update: programbench uses v6 images by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/867

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.1...v2.4.2

What's Changed

• fix: kill local process group on timeout by @Swing-G in https://github.com/SWE-agent/mini-swe-agent/pull/865

New Contributors

• @Swing-G made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/865
• maaz139 for reporting the timeout bug

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.4.0...v2.4.1

What's Changed

• Change(swe-bench configs): Drop T=0 default in swe-bench by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/861
• Enh: Expose finish_reason to format_error_template by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/862
• feat(inspector): display reasoning_content with toggle and hot-reload by @AlexanderDefuria in https://github.com/SWE-agent/mini-swe-agent/pull/815
• Enh: Cap consecutive format errors by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/863

New Contributors

• @AlexanderDefuria made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/815

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.3.1...v2.4.0

What's Changed

• Fix(swebench): source ~/.bashrc via BASH_ENV to activate testbed env by @seek-far in https://github.com/SWE-agent/mini-swe-agent/pull/858
• fix: persist full model response in extra dict on FormatError (#784) by @VittoriaLanzo in https://github.com/SWE-agent/mini-swe-agent/pull/821

New Contributors

• @VittoriaLanzo made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/821
• @Entrpi made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/850
• @seek-far made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/858

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.3.0...v2.3.1

What's Changed

The main feature is compatibility with ProgramBench, a new and ultra-challenging software benchmark.

Fixes

• fix: add exist_ok=True to mkdir in BubblewrapEnvironment by @hobostay in https://github.com/SWE-agent/mini-swe-agent/pull/802
• Fix/cost limit zero by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/825
• fix: add wall-clock time limit to properly kill agents by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/832

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.8...v2.3.0

[!WARNING] The dependency litellm in versions 1.82.7 and 1.82.8 was compromised. If you updated/installed litellm in the last few overs SSH and API keys might have been compromised.

The compromised versions should have been removed from pypi, but the PR below also excludes them in the dependencies.

• security: exclude compromised litellm versions 1.82.7 and 1.82.8 by @michalcichon in https://github.com/SWE-agent/mini-swe-agent/pull/794

New Contributors

• @michalcichon made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/794

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.7...v2.2.8

What's Changed

Mostly fixes to ContreeEnvironment:

• Fixed ContreeEnvironment for new contree-sdk version by @delphikettle in https://github.com/SWE-agent/mini-swe-agent/pull/765
• ContreeEnvironment Fixed double command execution + tests + docs fixes by @delphikettle in https://github.com/SWE-agent/mini-swe-agent/pull/777
• Enh: Generalize content string representation by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/783

New Contributors

• @hobostay made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/774

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.6...v2.2.7

Thanks to @AlienKevin for fixing a bug in the exception handling with the SwerexModalEnvironment

• fix(modal): Add missing exception_info to SwerexModalEnvironment output by @AlienKevin in https://github.com/SWE-agent/mini-swe-agent/pull/764

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.5...v2.2.6

Bugfix release fixing issues with FormatErrors especially for weaker/smaller models that make toolcalling mistakes, and a rare issue in the inspector that could hang if encountering ANSI-sequences in trajectories.

• Fix: check if args is a dict to avoid exceptions fro… by @RobinChiu in https://github.com/SWE-agent/mini-swe-agent/pull/760
• fix(inspector): Handle ANSI escape sequences and null bytes in content by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/761

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.4...v2.2.5

What's Changed

Fix a bug that caused an exception about missing default model name if setup is invoked in the first mini invocation right after installation.

• Fix: Reload config after setup by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/753

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.3...v2.2.4

Minor bug fix.

• Fixed a bug inside ContreeEnvironmentConfig.get_tag_by_image_url by @delphikettle in https://github.com/SWE-agent/mini-swe-agent/pull/751

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.2...v2.2.3

What's Changed

• Enh: Make 'actions' available in all format error templates by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/748
• Enh: Improve startup time of mini by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/749

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.1...v2.2.2

What's Changed

Small patch fixing SwerexModalEnvironment. Thanks to @arbipher!

• Fix: Update SwerexModalEnvironment for v2 Environment protocol by @arbipher in https://github.com/SWE-agent/mini-swe-agent/pull/747

New Contributors

• @arbipher made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/747

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.2.0...v2.2.1

What's Changed

Main feature is the new ContreeEnvironment develoepd by Nebius. Thanks @delphikettle for implementing it! You should now also be able to run on swe re-bench with the main swebench.py script.

• small typo in SWE-bench prompt by @a-scarlett in https://github.com/SWE-agent/mini-swe-agent/pull/746
• Added ContreeEnvironment by @delphikettle in https://github.com/SWE-agent/mini-swe-agent/pull/628

We've also finally updated the official SWE-bench leaderboard with the results from mini-swe-agent v2:

You can find more information in this tweet.

New Contributors

Welcome ❤️

• @a-scarlett made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/746
• @delphikettle made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/628

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.1.0...v2.2.0

This minor release improves the UX of the interactive agent (i.e., what is called when you run mini). We no longer prompt with the multiline prompt at the end of the agent run. You can still submit a multiline new task by entering /m first.

• Enh(interactive): Better behavior on exit by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/741
• Fix: Count of actions to confirm by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/742

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v2.0.0...v2.1.0

[!warning] This is the first full release of mini-swe-agent v2. This includes lots of breaking changes.

This is the largest new release since the launch, focusing on supporting more diverse models and switching to toolcalls by default (as opposed to manual parsing of actions).

What's new?

• Tool calls: Native tool calling API is now the support, text-based action parsing is still available
• Multimodal input: Support for images and other content types
• More flexible model support in general

What do I need to change?

I only use the mini CLI with default configs

No changes needed.

I use custom configs

You might need to move some config keys. See Config changes.

I parse/analyze trajectories

Some metadata fields moved to extra. See Trajectory format.

I use the python bindings or built custom subclasses

You will need to refactor. An agent can refactor your code based on the instructions below.

Full migration guide

Read the full migration guide.

Changes compared to v2.0.0a3:

• Enable parallel tool calls and use plain text observation template by @aorwall in https://github.com/SWE-agent/mini-swe-agent/pull/722
• Enh: Only ask for task if unset by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/738

Changelog compared to last v1

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.17.3...v2.0.0

[!warning] v2 coming on Monday! More flexibility & performance, but backwards breaking changes. Pin your dependency to v1 to avoid interruption and then check the migration guide.

This small patch release fixes a clash in the -c options for the swebench runner.

Thanks to @bofia for the fix in https://github.com/SWE-agent/mini-swe-agent/pull/720

[!note] v1.x releases are now based on the v1 branch. The v2 branch will soon become the main branch.

[!warning] v2 coming soon! More flexibility & performance, but backwards breaking changes. Pin your dependency to v1 to avoid interruption.

This is (hopefully) one of the last mini v1 releases, mostly fixing two bugs:

• Failed submission commands in swebench still submit (https://github.com/SWE-agent/mini-swe-agent/pull/691), thanks to @aorwall
• mini executable ignored environment config in custom config files (https://github.com/SWE-agent/mini-swe-agent/pull/700), thanks to @IKACE for the report

[!note] v1.x releases are now based on the v1 branch. The v2 branch will soon become the main branch.

This release adds important fixes.

Fixed

• fix: Reraise original exception when retries are exhausted by @aorwall in https://github.com/SWE-agent/mini-swe-agent/pull/638
• fix: add truncation to timeout_template to prevent context overflow by @aorwall in https://github.com/SWE-agent/mini-swe-agent/pull/639
• Fix: Remove timestamp field from OpenAI Response API requests by @aorwall in https://github.com/SWE-agent/mini-swe-agent/pull/637

New Contributors

• @aorwall made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/638

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.17.2...v1.17.3

Improved

• Enh: Keep batch progress main bar always on top by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/634

Fixed

• Fix: Avoid passing extra message information to model by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/633

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.17.1...v1.17.2

Minor fixes.

Fixes

• Fix: Make test_whitelist_actions_bypass_confirmation more robust by @Chesars in https://github.com/SWE-agent/mini-swe-agent/pull/620
• Fix(roulette config): Undefined variable by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/624
• Fix: Explicitly setting model name for litellm cost calc by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/626

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.17.0...v1.17.1

Improved

• Enh: Include timestamps & parsed actions in traj output by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/618

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.16.0...v1.17.0

This release adds more options to support models, in particular to use the responses API from OpenAI that is necessary to use Codex models, among others.

Added

• feat: make action parsing regex configurable by @Chesars in https://github.com/SWE-agent/mini-swe-agent/pull/564
• feat: add Requesty model provider support by @Thibault00 in https://github.com/SWE-agent/mini-swe-agent/pull/536
• feat: responses api by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/604
• Feat(models): Add portkey responses API by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/610

New Contributors

• @RicardoDominguez made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/551
• @Thibault00 made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/536
• @nihaljn made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/608

Welcome on board! ❤️

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.15.0...v1.16.0

Improved

This version introduces a switch to ignore errors about cost tracking: Set cost_tracking: "ignore_errors" in your agent config or export MSWEA_COST_TRACKING="ignore_errors" globally to ignore errors about missing cost information for local or free models. This is an easier alternative to setting up custom model registries etc. Thank you to @Chesars for initial work on this.

See https://github.com/SWE-agent/mini-swe-agent/pull/593, https://github.com/SWE-agent/mini-swe-agent/pull/594, https://github.com/SWE-agent/mini-swe-agent/pull/596

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.14.4...v1.15.0

Fixed

• Fix: Type error with portkey cost tracking workaround by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/580

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.14.3...v1.14.4

Small release containing fix for mini --cost-limit 0 bug spotted by @Chesars (thank you for your continued contributions)

Improved

• Enh: Configure number of retries for model by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/576

Fix

• Fix: --cost-limit flag doesn't work with 0 value due to Python falsy check by @Chesars in https://github.com/SWE-agent/mini-swe-agent/pull/575

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.14.2...v1.14.3

Portkey returns inconsistent token counts for anthropic models as soon as tokens are cached that can even lead to negative calculated costs. This is a workaround. This only affected you if you were using --model-class portkey or similar.

Fixed

• Fix: Portkey anthropic cost calculation bug workaround by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/573

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.14.1...v1.14.2

Quick but important fixup to previous release (the set_cache_control method had been accidentally overwritten with a nonsensical implementation).

Fixed

• Fix: Broken set_cache_control not clearing cache points by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/572

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.14.0...v1.14.1

This release introduces changes to how we handle Anthropic models, in particular with respect to caching. These are mostly internal changes, it is unlikely that you need to change anything on your end. The advantage is that you can now use Anthropic models also with other "backends" like portkey rather than having to use the litellm class.

Here's the breakdown of the changes:

1. Simpler cache breakpoint setting: We only apply one cache breakpoint at the end
2. If no model class is specified Anthropic models are also handled by litellm
3. All model classes apply the cache breakpoints automatically if the model name suggest an Anthropic model, unless the set_cache_control option is explicitly set to None
4. Anthropic key rotation is deprecated (as are some minor internal methods)

Changed

• Change: Anthropic caching applied at end; deprecate cache-related options by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/570
• Change: Use litellm class for anthropic; cache control support by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/571

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.13.4...v1.14.0

This patch release mostly fixes a bug in the command palette in visual (textual) mode reported & fixed by @Chesars

Improved

• Update default model to claude-sonnet-4-5-20250929 by @Chesars in https://github.com/SWE-agent/mini-swe-agent/pull/555

Fixed

• Fix: Ensure we use anthropic/ prefix for litellm by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/553
• Fix: Command palette crashes with TypeError in visual mode by @Chesars in https://github.com/SWE-agent/mini-swe-agent/pull/566

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.13.3...v1.13.4

This fixes a bug reported by @lamanchy affecting Windows/WSL where a system property (processor type) referenced in the default prompt template isn't present.

Fixed

• Fix: Remove 'processor' in template, might be undefined by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/548

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.13.2...v1.13.3

Fix un-incremented version of last release.

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.13.1...v1.13.2

Very minor improvement for portkey support.

Improved

• Enh: Allow to override litellm model name in portkey models by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/541

New Contributors

• @Chesars made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/537

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.13.0...v1.13.1

Adding portkey support.

Added

• Feat: Add portkey support by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/534

Improved

• Enh: Print agent config location in run scripts by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/532

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.12.0...v1.13.0

@riazarbi contributed an improvement to the action parser after observing that Qwen3-Coder 30B frequently adds extra whitespace before the linebreak after ```bash in codeblocks, causing mini to not recognize the action. The new regular expression tolerates whitespace before the linebreak.

Improved

• Handle whitespaces in bash command line by @riazarbi in https://github.com/SWE-agent/mini-swe-agent/pull/530
• Enh: Add a timeout argument as an override for env.execute by @SumanthRH in https://github.com/SWE-agent/mini-swe-agent/pull/517

New Contributors

• @SumanthRH made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/517
• @riazarbi made their first contribution in https://github.com/SWE-agent/mini-swe-agent/pull/530

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.11.1...v1.12.0

Fix for serialization errors introduced in v1.11.0

Fixed

• Fix: Response serialization error by @klieret in https://github.com/SWE-agent/mini-swe-agent/pull/519

Full Changelog: https://github.com/SWE-agent/mini-swe-agent/compare/v1.11.0...v1.11.1