DeerFlow 2.0: agent harness for research and coding

At a glance

Type

Framework

Workflow level

Workflow unspecified / not applicable

Disciplines

Computer science

Tasks

Knowledge synthesisLiterature researchResearch codingWorkflow orchestration

Entry updated

2026-10-17

Evidence & verification

Source level

Repository-documented
Grounded in a maintainer repository/README. Availability, stars and forks do not establish research benefit or correctness.

Outcome verification

Not empirically evaluated here
A documented resource or intended workflow, not a demonstrated academic outcome. No empirical benefit is established by its inclusion.

Sources checked

2026-10-17 (UTC)

Selected README/About, configuration, version and security passages reviewed from saved 2026-10-17 HTML; no code execution or exhaustive documentation review.

Source-checked does not mean independently reproduced. See the Method page for how these labels are assigned.

Sources (1)

Research applicability

Relevant as reusable infrastructure for agentic literature and coding workflows, and as a concrete comparator when evaluating research-agent architectures.

Getting started

The README documents cloning the repository and running make setup to choose a provider and safety preferences. Editorial suggestion, not exercised: begin in an isolated trusted environment with non-sensitive data and inspect every output before reuse.

Maintenance

The README says active development moved to 2.0, a ground-up rewrite sharing no code with v1. No maintenance guarantee or security audit.

License

MIT licence indicated on the repository page ; confirm terms before redistribution.

Workflow & human oversight

General orchestration framework; workflow level is unspecified until a concrete research workflow and oversight boundaries are defined.

Workflow level is not an evidence score; higher is not better evidence.

Approach

The project describes orchestration of subagents, memory, sandboxes and extensible skills for research and coding tasks.

Limitations

Selected repository-documentation review only; no code audit, installation or outcome reproduction. Version 2.0 differs from the original deep-research framework. Code execution and file access require restricted deployment; provider costs/privacy were not assessed.

Reproducibility

Source and configuration are public, but no run was attempted here; functional reproducibility is unverified.

Adoption requirements

Configure a supported LLM provider and execution/safety settings using the documented setup wizard. Editorial safeguards: assess provider privacy, restrict code/file access and review costs before supplying research data.

Comments

No comments yet.

Add a comment

Comments are held for moderation before appearing. Critical or corrective comments are welcome — especially evidence that an entry overstates what has been verified.