Skip to main content
hc
Submit
toolsSource-backed
mini-SWE-agent logo

mini-SWE-agent

MIT-licensed command-line software-engineering agent for local coding tasks, GitHub issue fixing, trajectory inspection, and SWE-bench style evaluation.

by SWE-agent · submitted by oktofeesh1·added 2026-06-03·
CLI
HarnessCLI
Review first — review before installing

Open the source and read safety notes before installing.

Citation facts

Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.

Canonical URL
https://heyclau.de/entry/tools/mini-swe-agent
Source URLs
https://mini-swe-agent.com/latest/, https://github.com/SWE-agent/mini-swe-agent, https://mini-swe-agent.com/
Brand
mini-SWE-agent
Brand domain
mini-swe-agent.com
Brand asset source
brandfetch
Safety notes
mini-SWE-agent is designed around language-model-generated terminal actions, so operators should review proposed actions before allowing changes in important repositories., The default CLI mode asks for confirmation before each proposed action, while automatic execution mode runs model-proposed actions without confirmation and should be limited to disposable or sandboxed environments., The project intentionally keeps the agent loop small and bash-oriented; that simplicity makes behavior easier to inspect, but it does not make actions safe, correct, authorized, or reversible., Generated patches, dependency changes, tests, issue updates, and benchmark runs still need human review before they affect protected branches, production systems, customer data, or shared infrastructure., Sandboxing choices such as local folders, containers, and isolated environments should be reviewed for mounted files, network access, dependency caches, and private access exposure., SWE-bench or other benchmark scores are useful evaluation signals, not proof that the agent is appropriate for a specific repository, policy boundary, or production workflow.
Privacy notes
Prompts, issue text, repository snippets, diffs, command output, test logs, and error traces may be sent to the configured model provider, gateway, or local model runtime., The CLI documentation says mini saves the full history of the last run to the global configuration directory, which can include local paths, source excerpts, model messages, and command output., Trajectory browser files, batch output, benchmark artifacts, and saved histories should be treated as potentially sensitive development records with retention and redaction rules., Local sandbox or container configuration can expose mounted source trees, dependency caches, environment settings, and private project access if the operator grants broad access., Teams using mini-SWE-agent for issue fixing should document who can access trajectories, logs, model-provider records, and any generated patches before sharing outputs.
Author
SWE-agent
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-03

Decision playbook

Review trust signals before you adopt

Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.

Compare context
Selected

0

Current score

78

Baseline

—

Delta

No baseline selected

No major trust-signal divergence detected in the current selection.

Source and provenance checks

Complete

Confirm ownership and provenance before trusting install instructions.

  • Source link availableRequired

    Open the canonical repository and verify ownership.

    Done

  • Source provenance statusRequired

    Marked as source-backed.

    Done

  • Metadata reviewed

    Registry metadata indicates a reviewed listing.

    Done

Safety and privacy checks

Complete

Validate risk disclosures before installation or API wiring.

  • Safety notes presentRequired

    Review the listed safety guidance before running commands.

    Done

  • Privacy notes presentRequired

    Review data handling notes before connecting accounts or secrets.

    Done

  • Trust level risk gateRequired

    Trust level does not block evaluation.

    Done

Package and install checks

Needs review

Check package metadata and artifact integrity signals.

  • Install payload available

    Install or copy payload is available for review.

    Done

  • Package verification flag

    No package verification flag provided.

    Pending

  • Checksum metadata

    No checksum provided for downloaded artifact.

    Pending

Compare-driven decision checks

Needs review

Use compare context to validate trade-offs before adoption.

  • Compare tray has multiple entries

    Add at least one more entry to compare trust differences.

    Pending

  • Baseline comparison available

    No baseline peer selected yet.

    Pending

  • Diverging trust signals identified

    No major trust-signal divergence found.

    Pending

Setup at a glance

Copy & paste

Copy-ready — paste the snippet to get started.

Install command

Not provided

Config snippet

Not provided

Copy snippet

Provided

Prerequisites

5 to clear

Platforms

1 listed

Install type

Copy & paste

Adoption plan

Balanced adoption plan

Current risk score 16/100. Use staged verification before broader rollout.

Risk 16

Pre-adoption checks

Validate source and review signals before any execution.

  • Confirm source provenanceRequired

    Source URL/provenance metadata is present.

    Done

  • Confirm metadata review state

    Listing has review metadata.

    Done

  • Verify install payload

    Install/config payload exists and can be inspected.

    Done

Security checks

Confirm safety, privacy, and package integrity signals.

  • Review safety notesRequired

    Safety notes are present.

    Done

  • Review privacy notesRequired

    Privacy notes are present.

    Done

  • Verify package integrity metadata

    No package verification/checksum metadata.

    Pending

Rollout

Adopt in controlled steps based on the selected plan.

  • Run in isolated sandbox firstRequired

    Use a constrained sandbox and observe behavior across multiple tasks.

    Pending

  • Roll out graduallyRequired

    Roll out to a small cohort before wider usage.

    Pending

  • Set monitoring and fallback

    Define rollback path and monitor errors after adoption.

    Pending

Evidence readiness

Evidence readiness matrix · balanced

Required evidence gates are covered (5/6 signals complete).

Risk 15

Source provenance

Present

Source repository/provenance is listed.

Required in this preset

Metadata review

Present

Review metadata is present.

Required in this preset

Safety notes

Present

Safety notes are present.

Required in this preset

Privacy notes

Present

Privacy notes are present.

Optional in this preset

Package integrity

Missing

Package integrity metadata is missing.

Optional in this preset

Install payload

Present

Install payload is available.

Required in this preset

Required evidence gates are covered for this preset.

Decision timeline

Decision timeline · balanced

5/6 steps complete with no blocking gaps for this preset.

Risk 14

triage

Confirm source provenanceRequired

Source/provenance metadata is available.

Done

triage

Check metadata review statusRequired

Review metadata is available.

Done

verify

Review safety notesRequired

Safety notes are available.

Done

verify

Review privacy notes

Privacy notes are available.

Done

verify

Validate package integrity metadata

Package integrity metadata is missing.

Pending

rollout

Verify install payload and commandsRequired

Install payload is available.

Done

No required blockers for this timeline preset.

Prerequisite readiness

Prerequisite readiness

5 prerequisites to line up before setup. Have accounts and credentials ready first. Includes a review or approval gate.

0/5 ready
Account & credentials1Permissions & scopes1Network & hosting1Review & approval1General1

Safety & privacy surface

Safety & privacy surface

6 safety and 5 privacy notes across 6 risk areas. Review closely: permissions & scopes, network access, third-party handling.

6 areas
  • SafetyGeneralmini-SWE-agent is designed around language-model-generated terminal actions, so operators should review proposed actions before allowing changes in important repositories.
  • SafetyExecution & processesThe default CLI mode asks for confirmation before each proposed action, while automatic execution mode runs model-proposed actions without confirmation and should be limited to disposable or sandboxed environments.
  • SafetyPermissions & scopesThe project intentionally keeps the agent loop small and bash-oriented; that simplicity makes behavior easier to inspect, but it does not make actions safe, correct, authorized, or reversible.
  • SafetyExecution & processesGenerated patches, dependency changes, tests, issue updates, and benchmark runs still need human review before they affect protected branches, production systems, customer data, or shared infrastructure.
  • SafetyNetwork accessSandboxing choices such as local folders, containers, and isolated environments should be reviewed for mounted files, network access, dependency caches, and private access exposure.
  • SafetyGeneralSWE-bench or other benchmark scores are useful evaluation signals, not proof that the agent is appropriate for a specific repository, policy boundary, or production workflow.
  • PrivacyThird-party handlingPrompts, issue text, repository snippets, diffs, command output, test logs, and error traces may be sent to the configured model provider, gateway, or local model runtime.
  • PrivacyLocal filesThe CLI documentation says mini saves the full history of the last run to the global configuration directory, which can include local paths, source excerpts, model messages, and command output.
  • PrivacyLocal filesTrajectory browser files, batch output, benchmark artifacts, and saved histories should be treated as potentially sensitive development records with retention and redaction rules.
  • PrivacyPermissions & scopesLocal sandbox or container configuration can expose mounted source trees, dependency caches, environment settings, and private project access if the operator grants broad access.
  • PrivacyThird-party handlingTeams using mini-SWE-agent for issue fixing should document who can access trajectories, logs, model-provider records, and any generated patches before sharing outputs.

Disclosure: editorial

Safety notes

  • mini-SWE-agent is designed around language-model-generated terminal actions, so operators should review proposed actions before allowing changes in important repositories.
  • The default CLI mode asks for confirmation before each proposed action, while automatic execution mode runs model-proposed actions without confirmation and should be limited to disposable or sandboxed environments.
  • The project intentionally keeps the agent loop small and bash-oriented; that simplicity makes behavior easier to inspect, but it does not make actions safe, correct, authorized, or reversible.
  • Generated patches, dependency changes, tests, issue updates, and benchmark runs still need human review before they affect protected branches, production systems, customer data, or shared infrastructure.
  • Sandboxing choices such as local folders, containers, and isolated environments should be reviewed for mounted files, network access, dependency caches, and private access exposure.
  • SWE-bench or other benchmark scores are useful evaluation signals, not proof that the agent is appropriate for a specific repository, policy boundary, or production workflow.

Privacy notes

  • Prompts, issue text, repository snippets, diffs, command output, test logs, and error traces may be sent to the configured model provider, gateway, or local model runtime.
  • The CLI documentation says mini saves the full history of the last run to the global configuration directory, which can include local paths, source excerpts, model messages, and command output.
  • Trajectory browser files, batch output, benchmark artifacts, and saved histories should be treated as potentially sensitive development records with retention and redaction rules.
  • Local sandbox or container configuration can expose mounted source trees, dependency caches, environment settings, and private project access if the operator grants broad access.
  • Teams using mini-SWE-agent for issue fixing should document who can access trajectories, logs, model-provider records, and any generated patches before sharing outputs.

Prerequisites

  • Python environment, package manager, and supported local shell environment for running the mini-SWE-agent CLI, batch workflows, or Python bindings.
  • Approved model route, provider account, local model setup, or gateway configuration prepared outside the repository where the agent will run.
  • Disposable checkout, sandbox, container, or tightly scoped working tree for tasks that may modify source files, dependencies, tests, or generated artifacts.
  • Clear task prompt, target branch, test command, rollback plan, and reviewer ownership before asking the agent to work on a real issue or codebase.
  • Retention plan for saved histories, trajectories, command output, benchmark artifacts, and any local logs produced while the agent is running.

Schema details

Install type
copy
Troubleshooting
No
Source repository stats
Scope
Source repo
Tool listing metadata
Pricing
open-source
Disclosure
editorial
Application category
DeveloperApplication
Operating system
macOS, Windows, Linux
Full copyable content
## Editorial notes

mini-SWE-agent is useful when Claude-adjacent teams want a small, auditable coding-agent CLI for local tasks, issue fixing, research baselines, or SWE-bench style evaluation. Its center of gravity is intentionally minimal: a compact agent loop, local environment support, terminal-action based operation, confirm and human modes, optional automatic execution, trajectory inspection, batch inference, and Python extension points.

This is distinct from existing coding-assistant and agent-framework entries. Claude Code, Cline, Roo Code, Aider, Continue, Cursor, Windsurf, Replit Agent, and Sourcegraph Cody focus on broader IDE, editor, or coding-assistant workflows. LangGraph, Pydantic AI, Agno, CrewAI, and AutoGen focus on agent framework construction. mini-SWE-agent is narrower: a research-backed, minimal software-engineering agent that uses simple local control flow for CLI coding tasks, issue repair, trajectories, and benchmark runs.

## Source notes

- The official README describes mini-SWE-agent as a minimal AI software-engineering agent built by the Princeton and Stanford team behind SWE-bench and SWE-agent.
- The README says mini is widely adopted, intentionally minimal, MIT licensed, compatible with many model routes, and designed for local environments, containers, and other sandboxing approaches.
- The README says mini does not use custom tools beyond bash, does not require model tool-calling interfaces, executes independent actions, and keeps a linear history for debugging and fine-tuning.
- The README recommends mini-SWE-agent as the default choice when users want a quick command-line tool, simple control flow, sandboxing, benchmark evaluation, or a fine-tuning/reinforcement-learning baseline.
- The official quick-start documentation covers package-manager based installation paths, CLI and extra utility entry points, source installation, model setup, local model guidance, and example coding prompts.
- The official CLI documentation describes `mini` as a REPL-style interactive command-line interface with confirm, human, and automatic execution modes; confirm mode is the default, and the last run history is saved locally.

## Duplicate check

Checked current `content/tools/`, `content/mcp/`, agents, hooks, rules, skills, commands, guides, open pull requests, live issue state, and repository-wide content for `mini-SWE-agent`, `mini swe agent`, `mini-swe`, `SWE-agent/mini-swe-agent`, `mini-swe-agent.com`, `SWE-agent`, `SWE-bench agent`, `coding agent CLI`, and `trajectory browser`. Existing Claude Code, Cline, Roo Code, Aider, Continue, Cursor, Windsurf, Replit Agent, Sourcegraph Cody, LangGraph, Pydantic AI, Agno, CrewAI, and AutoGen entries cover adjacent coding assistant or framework workflows, but no dedicated mini-SWE-agent tools entry, mini-SWE-agent source URL duplicate, or open duplicate PR was found.

## Disclosure

Editorial listing. No paid placement or affiliate link is used.

About this resource

Editorial notes

mini-SWE-agent is useful when Claude-adjacent teams want a small, auditable coding-agent CLI for local tasks, issue fixing, research baselines, or SWE-bench style evaluation. Its center of gravity is intentionally minimal: a compact agent loop, local environment support, terminal-action based operation, confirm and human modes, optional automatic execution, trajectory inspection, batch inference, and Python extension points.

This is distinct from existing coding-assistant and agent-framework entries. Claude Code, Cline, Roo Code, Aider, Continue, Cursor, Windsurf, Replit Agent, and Sourcegraph Cody focus on broader IDE, editor, or coding-assistant workflows. LangGraph, Pydantic AI, Agno, CrewAI, and AutoGen focus on agent framework construction. mini-SWE-agent is narrower: a research-backed, minimal software-engineering agent that uses simple local control flow for CLI coding tasks, issue repair, trajectories, and benchmark runs.

Source notes

  • The official README describes mini-SWE-agent as a minimal AI software-engineering agent built by the Princeton and Stanford team behind SWE-bench and SWE-agent.
  • The README says mini is widely adopted, intentionally minimal, MIT licensed, compatible with many model routes, and designed for local environments, containers, and other sandboxing approaches.
  • The README says mini does not use custom tools beyond bash, does not require model tool-calling interfaces, executes independent actions, and keeps a linear history for debugging and fine-tuning.
  • The README recommends mini-SWE-agent as the default choice when users want a quick command-line tool, simple control flow, sandboxing, benchmark evaluation, or a fine-tuning/reinforcement-learning baseline.
  • The official quick-start documentation covers package-manager based installation paths, CLI and extra utility entry points, source installation, model setup, local model guidance, and example coding prompts.
  • The official CLI documentation describes mini as a REPL-style interactive command-line interface with confirm, human, and automatic execution modes; confirm mode is the default, and the last run history is saved locally.

Duplicate check

Checked current content/tools/, content/mcp/, agents, hooks, rules, skills, commands, guides, open pull requests, live issue state, and repository-wide content for mini-SWE-agent, mini swe agent, mini-swe, SWE-agent/mini-swe-agent, mini-swe-agent.com, SWE-agent, SWE-bench agent, coding agent CLI, and trajectory browser. Existing Claude Code, Cline, Roo Code, Aider, Continue, Cursor, Windsurf, Replit Agent, Sourcegraph Cody, LangGraph, Pydantic AI, Agno, CrewAI, and AutoGen entries cover adjacent coding assistant or framework workflows, but no dedicated mini-SWE-agent tools entry, mini-SWE-agent source URL duplicate, or open duplicate PR was found.

Disclosure

Editorial listing. No paid placement or affiliate link is used.

#coding-agent#cli#open-source

Source citations

Add this badge to your README

Show that mini-SWE-agent is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.

Listed on HeyClaude
[![Listed on HeyClaude](https://heyclau.de/badge/tools/mini-swe-agent.svg)](https://heyclau.de/entry/tools/mini-swe-agent)

How it compares

mini-SWE-agent side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.

1 trust signal differ across this comparison (Submitter).

Field
mini-SWE-agent logomini-SWE-agent

MIT-licensed command-line software-engineering agent for local coding tasks, GitHub issue fixing, trajectory inspection, and SWE-bench style evaluation.

Open dossier
Cline logoCline

Open-source autonomous coding agent extension for planning, editing, running commands, and using tools from VS Code.

Open dossier
Open SWE

Open-source framework for building internal coding agents that accept tasks via Slack, Linear, or GitHub, execute code changes in isolated cloud sandboxes, and open draft pull requests automatically.

Open dossier
OpenAI logoOpenAI Codex CLI

Lightweight, open-source coding agent from OpenAI that runs in your terminal, edits and runs code locally with configurable sandbox and approval modes, and authenticates with a ChatGPT plan or API key.

Open dossier
Next steps
Open dossierAPI JSONOpen LLMOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMOpen sourceNewsletterClaim listing
Trust
Review statusReviewedMaintainer reviewedReviewedMaintainer reviewedReviewedMaintainer reviewedReviewedMaintainer reviewed
Package trustPackage not verifiedPackage not verifiedPackage not verifiedPackage not verified
Source provenanceSource-backedSource-backedSource-backedSource-backed
SubmitterDiffersoktofeesh1—JPette1783JPette1783
Install riskReview firstReview firstReview firstReview first
Notes Safety ✓ Privacy ✓ Safety · Privacy · Safety ✓ Privacy ✓ Safety ✓ Privacy ✓
Brandmini-SWE-agent logomini-SWE-agentCline logoCline—OpenAI logoOpenAI
Categorytoolstoolstoolstools
SourceSource-backedSource-backedSource-backedSource-backed
AuthorSWE-agentClineLangChainOpenAI
Added2026-06-032026-04-272026-06-052026-06-05
Platforms
CLI
CLI
CLI
CLI
Harness
CLI
CLI
CLI
CLI
Source repo————
Safety notes✓mini-SWE-agent is designed around language-model-generated terminal actions, so operators should review proposed actions before allowing changes in important repositories. The default CLI mode asks for confirmation before each proposed action, while automatic execution mode runs model-proposed actions without confirmation and should be limited to disposable or sandboxed environments. The project intentionally keeps the agent loop small and bash-oriented; that simplicity makes behavior easier to inspect, but it does not make actions safe, correct, authorized, or reversible. Generated patches, dependency changes, tests, issue updates, and benchmark runs still need human review before they affect protected branches, production systems, customer data, or shared infrastructure. Sandboxing choices such as local folders, containers, and isolated environments should be reviewed for mounted files, network access, dependency caches, and private access exposure. SWE-bench or other benchmark scores are useful evaluation signals, not proof that the agent is appropriate for a specific repository, policy boundary, or production workflow.— missing✓Each task runs in an isolated cloud Linux sandbox (Modal, Daytona, Runloop, or LangSmith) to prevent production impact. The agent executes shell commands, file operations, web fetches, and HTTP requests inside the sandbox without confirmation prompts — review sandbox provider permissions before deployment. GitHub operations are performed through a GH_TOKEN proxy; scope token permissions to the minimum required repositories. Subagent orchestration can spawn parallel child agents — set appropriate step limits and monitor LangSmith traces to prevent runaway execution. AGENTS.md or CLAUDE.md at the repository root is injected into the system prompt; review this file to control agent behavior and conventions.✓Codex CLI runs as a local agent that can edit files and execute commands on your machine. It supports configurable sandbox and approval modes; keep approvals enabled and review proposed actions before allowing them. Lowering approval or sandbox restrictions lets the agent act more autonomously, increasing the risk of unintended changes. Review generated code and commands before running them against important repositories or systems.
Privacy notes✓Prompts, issue text, repository snippets, diffs, command output, test logs, and error traces may be sent to the configured model provider, gateway, or local model runtime. The CLI documentation says mini saves the full history of the last run to the global configuration directory, which can include local paths, source excerpts, model messages, and command output. Trajectory browser files, batch output, benchmark artifacts, and saved histories should be treated as potentially sensitive development records with retention and redaction rules. Local sandbox or container configuration can expose mounted source trees, dependency caches, environment settings, and private project access if the operator grants broad access. Teams using mini-SWE-agent for issue fixing should document who can access trajectories, logs, model-provider records, and any generated patches before sharing outputs.— missing✓Repository code, Linear issue history, and Slack thread history are sent to the configured model provider API. Sandbox providers (Modal, Daytona, Runloop, LangSmith) process task execution data according to their own privacy policies. LangSmith tracing, when enabled, logs full agent traces including tool inputs and outputs — configure retention and access controls in your LangSmith organization. GitHub OAuth tokens and model API keys should be stored as secrets and never committed to the repository.✓Code and prompts are sent to OpenAI to power the agent under your ChatGPT plan or API key. Data handling follows OpenAI's policies for the plan or API you authenticate with; review them before processing sensitive code. API keys and account credentials should be stored securely and never committed to source control.
Prerequisites
  • Python environment, package manager, and supported local shell environment for running the mini-SWE-agent CLI, batch workflows, or Python bindings.
  • Approved model route, provider account, local model setup, or gateway configuration prepared outside the repository where the agent will run.
  • Disposable checkout, sandbox, container, or tightly scoped working tree for tasks that may modify source files, dependencies, tests, or generated artifacts.
  • Clear task prompt, target branch, test command, rollback plan, and reviewer ownership before asking the agent to work on a real issue or codebase.
— none listed
  • GitHub account with OAuth access for repository operations.
  • A model API key (Anthropic, OpenAI, or compatible provider).
  • A LangSmith API key when using LangSmith as the sandbox provider.
  • Slack workspace, Linear workspace, or GitHub repository access for the desired trigger integrations.
  • A ChatGPT plan (Plus, Pro, Business, Edu, or Enterprise) for plan-based sign-in, or an OpenAI API key.
  • Node.js and npm for the npm install, or Homebrew for the cask install.
  • A terminal on macOS, Linux, or Windows.
Install————
Config————
Citations
ClaimUnclaimedUnclaimedUnclaimedUnclaimed
Open 4 picks in the interactive comparison tool

Related guides

Source-backed guides for putting this to work.

Adding Plugin Install Hints to Developer CLIs

Write a claude-code-hint tag to stderr under CLAUDECODE, and Claude Code prompts users once to install your official-marketplace plugin.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by JPette1783

Coming to Claude Code from ChatGPT

Get productive with Claude Code if you're coming from ChatGPT.

Added 8mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by JSONbored

Resuming, Branching, and Naming Claude Code Sessions

Resume, name, branch, and switch between Claude Code sessions with --continue, --resume, /rename, /branch, and the session picker.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by JPette1783

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.