Technology · Agentic → Coding Agents · Open Source · wiki:deep
SWE-agent
SWE-agent (Princeton) frames software engineering as an agent interacting with a controlled interface (ACI) optimized for language models resolving GitHub issues. It is influential research/OSS for agent–repo interaction design. Use as evaluation and design reference for coding agents; not by itself an enterprise OS runtime. SWE-agent takes a GitHub issue and tries to automatically fix it, using your LM of choice. It can also be employed for offensive cybersecurity or competitive coding challenges. [NeurIPS 2024] [!warning] Most of our current development effort is on mini-swe-agent, which has superseded SWE-agent. It matches the performance performance of SWE-agent, while being much simpler. See the FAQ for more details about the differences. Our general recommendation is to use mini-SWE-agent instead of SWE-agent going forward. SWE-agent enables your language model.
Why it matters here
In this catalog it is the Coding Agents option when you need to edit repositories and ship engineering changes. Keep SWE-agent on the observe/propose side of the loop: useful signals and drafts, with human or policy authority still required before anything irreversible runs.
How it works
SWE-agent uses an Agent-Computer Interface (ACI) tuned so LMs can resolve GitHub-style issues in a controlled environment.
- Provide an issue + repository snapshot.
- Agent acts through the ACI (edit, run, search).
- Tests/commands give feedback.
- Emit a patch candidate when checks pass.
Related: openhands · topics/06-self-coding
When to reach for it
- Use when: SWE-agent (Princeton) frames software engineering as an agent interacting with a controlled interface (ACI) optimized for language models resolving GitHub issues matches the bottleneck and you specifically need to edit repositories and ship engineering changes.
- Skip when: another pack already covers the need with less ops surface, or SWE-agent would become a second place people retype truth.
- Prefer instead: compare OpenHands before locking the cell.
Limits
- Ops: confirm license, hosting, and release cadence for SWE-agent (SWE-agent/SWE-agent) before relying on it in a live corridor.
- Integration: bad credentials, scopes, or schema drift produce empty retrievals or runaway tool/model spend without an obvious UI alarm.
- Scope: Coding Agents components support agents; they are not the company ledger or the approval authority.
What we checked
Claims below are backed by science sources on disk.
Primary / product role
SWE-agent/SWE-agent · MODERATE
“Run OpenHands, Claude Code, Codex, Gemini, or any ACP-compatible agent across local, remote, and cloud backends.”
Docs or contrast
Platform contrast · MODERATE
“The self-hosted developer control center for coding agents and automations.”
Additional source
SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering · MODERATE
“OpenHands Agent Canvas turns your coding agents into a self-hosted, always-on engineering team.”
Research inventory
9 tags · 9 out · 10 in · 3 artifacts · 1 gaps · 13 corpus docs
Catalog tags
- landscape.layer
- Agentic
- landscape.subcategory
- Coding Agents
- license_tag
- Open Source
- maps.dm
- absent
- maps.features
- 4
- one_liner
- Agentic
- review.depth
- science
- slug
- swe-agent
- title
- SWE-agent
Artifacts
- dm_map · absent
- features_map · present · technologies/swe-agent/features.md
- readme · present · technologies/swe-agent/README.md
Out · alternative_to
- aidertechnology
- claude-codetechnology
- continuetechnology
- open-interpretertechnology
- openhandstechnology
Out · maps_to
- job-codingfeature
- learning-adaptationfeature
- r2-06-self-codingfeature
- sandboxing-isolationfeature
In · alternative_to
- aidertechnology
- claude-codetechnology
- continuetechnology
- open-interpretertechnology
- openhandstechnology
In · in_stack
- agentic-coding-agentsstack · alternative_slot
In · maps_to
- job-codingfeature
- learning-adaptationfeature
- r2-06-self-codingfeature
- sandboxing-isolationfeature
Corpus tags
- category
- Agentic → Coding Agents
- dedication
- open-source
- feature
- job-codinglearning-adaptationr2-06-self-codingsandboxing-isolation
- kind
- map_edgetech_featurestech_quotetech_readme
- needs_deepen
- false
- quality
- ok
- slug
- swe-agent
- source_id
- openhands-contrastsweagent-githubsweagent-paper
- technology
- swe-agent
Corpus documents (13)
map_edge · 4
- swe-agent → job-coding
- swe-agent → learning-adaptation
- swe-agent → r2-06-self-coding
- swe-agent → sandboxing-isolation
tech_features · 1
- SWE-agent · features
tech_quote · 7
- SWE-agent · openhands-contrast
- SWE-agent · openhands-contrast
- SWE-agent · openhands-contrast
- SWE-agent · sweagent-github
- SWE-agent · sweagent-github
- SWE-agent · sweagent-paper
- SWE-agent · sweagent-paper
tech_readme · 1
- SWE-agent