Skip to main content

SOAT vs DeepSeek Harness

DeepSeek Harness is an agent framework (MIT, self-hosted). SOAT is agent infrastructure (Apache-2.0, self-hosted). Of the 15 capabilities the benchmark tracks, DeepSeek Harness covers 4 natively and SOAT 12.

At a glance​

DeepSeek HarnessSOAT
Typean agent frameworkagent infrastructure
LicenseMITApache-2.0
Deploymentself-hostedself-hosted
Native capabilities4 of 1512 of 15
Last verified2026-08-182026-08-14

Where each one leads​

  • DeepSeek Harness is native where SOAT is not: code execution.
  • SOAT is native where DeepSeek Harness is not: orchestration, knowledge & memory, identity & governance, observability, data & secrets, evaluation, multi-tenancy, declarative deployment, versioning & rollout.

What DeepSeek Harness is​

Plugin-based open-source agent framework from DeepSeek AI ("everything is a plugin") built on the Cordis composable plugin system, with a web UI and CLI for building and running custom agents. DeepSeek Harness documentation.

Capability by capability​

Agent runtime​

Agent loop, sessions, and generation lifecycle.

  • DeepSeek Harness: Native. Core agent loop and web UI for agent interaction. (source)
  • SOAT: Native. Agents, sessions, chats, and generations with full lifecycle tracking. (source)

Orchestration​

Multi-step pipelines, workflows, and triggers.

  • DeepSeek Harness: Partial. A built-in workflow family runs model-authored dynamic workflows over subagents, plus plan/schedule/goal packages; no declarative DAG with explicit steps, dependencies, or conditions. (source)
  • SOAT: Native. DAG orchestrations, workflow state machines, durable tasks, and triggers (manual, webhook, schedule). (source)

Knowledge & memory​

Documents, embeddings, semantic search, and recall.

  • DeepSeek Harness: Via plugin. No built-in document store, embeddings, or memory — provided by plugins. (source)
  • SOAT: Native. Documents with per-chunk embeddings, named memories, unified semantic search, and ingestion rules for PDFs, images, and audio. (source)

Tools & integration​

Tool calling, MCP, providers, and webhooks.

  • DeepSeek Harness: Native. The plugin system is the tool/extension surface; capabilities are added as Cordis plugins. (source)
  • SOAT: Native. HTTP, MCP, client, and agent-to-agent tools; model routes with failover; webhooks; any LLM provider. (source)

Identity & governance​

IAM, guardrails, approvals, quotas, and audit.

  • DeepSeek Harness: Absent. No multi-tenant IAM, policies, quotas, approvals, or audit log.
  • SOAT: Native. AWS-inspired IAM, actors, scoped API keys, guardrails, human approvals, quotas, and an append-only audit log. (source)

Observability​

Traces, activity, usage, and failure queues.

  • DeepSeek Harness: Partial. Runtime diagnostics, a per-session audit trail, and durable JSONL/SQLite session persistence with a query layer; no structured trace/usage metering model. (source)
  • SOAT: Native. Traces of every tool call and model response, activity feed, usage/cost tracking, and an exceptions queue. (source)

Data & secrets​

File storage and encrypted secrets.

  • DeepSeek Harness: Partial. Credential references resolved from env/local files (config carries references, not values) plus session storage and attachments; no managed encrypted secrets store or API-managed file storage. (source)
  • SOAT: Native. File storage over local/S3/GCS backends and encrypted secrets storage. (source)

Evaluation​

Datasets, scorers, and eval-gated improvement.

  • DeepSeek Harness: Absent. No built-in datasets, scorers, or eval runs.
  • SOAT: Native. Datasets, scorers, and scored runs; eval-gated promotion of agent versions. (source)

Channels​

Built-in gateways to messaging platforms and UIs.

  • DeepSeek Harness: Via plugin. Web UI is built in; messaging-platform gateways arrive as Cordis plugins. (source)
  • SOAT: Absent. Surfaces are REST, MCP, CLI, and SDK — no built-in gateways to messaging platforms like Slack, WhatsApp, or Telegram.

Code execution​

Sandboxed code or terminal runtimes for agents.

  • DeepSeek Harness: Native. First-party per-session process confinement (bwrap/Landlock/Seatbelt backends) and a code-execution family running in worker threads, with an E2B integration; worker threads are not a security boundary. (source)
  • SOAT: Absent. Tool types are HTTP, MCP, client, and agent-to-agent — no built-in sandboxed code or terminal runtime.

Human-in-the-loop​

Pausing a run for human approval, then resuming it.

  • DeepSeek Harness: Native. An approval package supplies a fail-closed approval service that gates each tool call under a per-session ask or never policy, plus a seam for free-form questions to the user. (source)
  • SOAT: Native. Approvals are a first-class human-decision queue with frozen evidence, and guardrails classify each tool call so risky actions pause for a decision. (source)

Multi-tenancy​

Projects or workspaces scoping resources and permissions.

  • DeepSeek Harness: Absent. Workspaces are local user directories holding session membership, and the identity package states its values do not represent an authenticated account, so no tenant-scoped permission model exists.
  • SOAT: Native. Projects are multi-tenant namespaces scoping every resource, with IAM policies and quotas enforced per project. (source)

Declarative deployment​

Templates that provision agent resources as one stack.

  • DeepSeek Harness: Partial. An agent preset declaratively composes a session's tools, persona, and prompt sections, but it is per-session composition with no plan/apply, dependency graph, or resource provisioning. (source)
  • SOAT: Native. Formations provision providers, memories, tools, agents, orchestrations, and webhooks from one template, resolving refs into a dependency graph with tracked operations. (source)

Versioning & rollout​

Agent versions, canary rollout, gated promotion.

  • DeepSeek Harness: Absent. No versioned agent definitions, aliases, canary rollout, or rollback appear in the packages or subsystem documentation.
  • SOAT: Native. Every config change is archived as an append-only agent version, with canary rollout to a slice of traffic and eval-gated promotion. (source)

Skill learning​

Agents authoring reusable skills from their own experience.

  • DeepSeek Harness: Absent. The skill family only discovers pre-authored skills through filesystem providers and exposes a read-only catalog and loader; there is no write path for the agent to author a skill.
  • SOAT: Absent. Memories and knowledge store retrievable context, but agents do not author or refine reusable skills from their own experience.

How this is rated​

Each rating comes from the public documentation linked beside it: Native is built in, Via plugin needs an add-on, Partial covers part of the capability, Absent is not offered. Every solution side by side is on the benchmark. A rating that is out of date is a bug: open an issue.