DeepSeek Harness Handbook

August 22, 2026 · View on GitHub

English · 简体中文 · 日本語 · 한국어 · Español

GitHub stars Content check License: Apache-2.0

Listed in Awesome DeepSeek Harness, the community-maintained DSH ecosystem directory.

DeepSeek Harness Handbook — operator guides, rc.8, Install Doctor, and Failure Router

The agent-first, English-canonical field guide to understanding, running, debugging, and extending DeepSeek Harness, with reviewed Simplified Chinese coverage and multilingual foundations. — Community-maintained and open-source.

115 canonical guides · rc.8 and rc.2 source coverage · primary-source links · runnable recovery paths

Unlike a command catalog, this handbook follows the complete Agent boundary: model routing, tools, approval, sandboxing, durable Sessions, plugins, MCP, ACP, and operator-visible failure recovery. Version-sensitive pages identify the source revision they were checked against.

Choose one path and get to evidence quickly:

RunDebugBuild
Five-minute quickstartInstall DoctorFirst plugin lab
CLI mapFailure RouterMCP guide
API cost boundaryCurrent field statusAgent runtime map

If one of these guides saves an incident or an afternoon, star the repository. Stars help the next Agent builder find a source-backed answer instead of another unverified command list.

Browse the visual field guides · Request a source-backed runbook · Subscribe to new guides · Read the changelog

DeepSeek Harness is more than a model wrapper. It is a composable agent runtime that connects model providers, tools, approval, sandboxing, durable sessions, subagents, and user interfaces through a plugin graph. This independent handbook explains those systems from the perspective of people building and operating agents.

The project is maintained by SandBase. It is not an official DeepSeek AI project.

Important

DeepSeek Harness is in developer preview and may introduce compatibility-breaking changes. Pages in this handbook name their verification date and link to primary sources. Pin the revision you deploy.

Start with your goal

I want to…Start here
Diagnose manual /compact ending with DeepSeek request aborted by callerManual compaction cancellation runbook
Fix ERR_PNPM_UNEXPECTED_STORE during a plugin updatepnpm store-identity recovery
Fix AbortSignal.any is not a function even though Node looks currentRuntime identity and offline recovery runbook
Recover Web stuck on Loading plugins in a pnpm source checkoutpnpm symlink boot guide
Separate Responses API full-history traffic, retry attempts, and SSE leaksResponses overload runbook
Run supported Codex hooks without assuming policy parityCodex hooks bridge guide
Install the Claude Code hooks bridge without duplicating the runtimeHooks bridge installation guide
Recover when a second core package copy breaks every tool callDuplicate core runtime recovery
Design session, model, JSON, and exit semantics for headless embeddingProgrammatic headless contract
Build rc.8 in an air-gapped environment without losing provenanceAir-gapped source-build guide
Interpret token estimates, provider usage, UI occupancy, and compaction pressureToken accounting map
Recover Session creation after editing a live Agent presetPreset generation recovery
Recover raw pinyin, kana, or jamo in the Web composerWeb IME composition runbook
Protect API keys from backups, same-UID tools, or an untrusted AgentCredential storage threat model
Recover a Session whose committed event sequence repeatsDuplicate committed seq runbook
Recover a Web composer stuck read-only after sending an imageImage-send admission runbook
Set and verify reasoning effort for a headless one-shot runHeadless reasoning-effort guide
Configure Bailian Token Plan without losing reasoning or model metadataBailian catalog-route runbook
Recover tools that repeatedly return Unknown or expired MCP sessionExpired MCP session loop runbook
Understand the shipped CLI, automate one task, or evaluate a community TUIDeepSeek Harness CLI map
Recover when the Windows folder picker crashes or truncates a Unicode pathWindows folder-picker crash and truncation guide
Connect to DeepSeek through an authorized proxy or enterprise CAProvider egress and TLS guide
Fix Web Search authentication when chat uses a custom gatewayCustom-gateway Web Search runbook
Add an MCP server or diagnose missing MCP toolsMCP preset and connection guide
Evaluate DeepSeek Harness as an ACP External Agent in an editorACP editor-integration boundary
Render an ACP permission request in a custom Web clientACP permission UI contract
Move Codex or Claude Code memory without losing provenance or isolationSessions, Skills, and long-term memory
Run Web, headless, ACP, or SDK processes concurrently without sharing a writable Session rootSingle-writer Session topology
Fix spawn bash ENOENT after moving or renaming a workspaceMoved-workspace recovery runbook
Upgrade an exact runtime and preserve a proven rollbackUpgrade and rollback guide
Detect when a community plugin replaces core Agent providersComposition-diff plugin audit
Choose an exact, reproducible installation topologyInstall DeepSeek Harness safely
Diagnose npx waiting at the DeepSeek Harness install promptnpx install-boundary runbook
Create, invoke, and debug a reusable SkillDeepSeek Harness Skills lab
Decide whether an instruction belongs in global, project, nested, local, or Skill scopeAGENTS.md scope and precedence map
Build, test, package, and install my first pluginFirst DeepSeek Harness plugin lab
Keep a plugin subprocess from freezing the Agent HostAsync subprocess tool guide
Discover and manage community plugins from the DSH Web UIDSH Plugin Store (GitHub)
Fix a plugin boot crash for missing @deepseek-ai/dsh-client-schema-formPlugin distribution-closure runbook
Fix additionalProperties, type-array, or oneOf tool schema errorsTool schema subset guide
Distinguish the official Agent runtime from same-name API wrappersOfficial DeepSeek Harness identity guide
Check current rc.7 boundaries and safer next actionsDeepSeek Harness rc.7 Field Status
Capture the package and source revision that actually ranDeepSeek Harness Version Evidence
Find the first broken runtime boundaryInteractive Failure Router
Keep the essential commands and checks in one tabDeepSeek Harness cheat sheet
Choose the right official runnable exampleOfficial examples map
Understand what DeepSeek Harness actually isDeepSeek Harness explained
Choose between DeepSeek Harness, Claude Code, and CodexSource-backed control-plane comparison
Read a Session log without dropping packed assistant outputSession log storage-format map
Recover when an invalid overlay prevents the profile from bootingInvalid-overlay recovery runbook
Run the Web UI safelyFive-minute quickstart
Use it from PythonPython SDK quickstart
Run one task in automation or CICLI and Headless Agent guide
Configure DeepSeek or another providerModel provider guide
Track the rc.2 Files-backed image-input releaseRelease and migration note
Prove DeepSeek chat and Web search cannot create unexpected chargesDeepSeek API cost-boundary runbook
Fix a context-window or token-budget errorClassify and recover context overflow
Understand why one model switch affects future AgentsMap session and deployment model state
Fix a plugin script corrupted around a dollar signRepair tapIndex replacement-string insertion
Fix an OpenAI-compatible gateway that rejects developerDiagnose system-message role compatibility
Fix CJK commands that stall persistent Bash for 300 secondsSeparate the locale/readline trap from PTY corruption
Fix plugin installation blocked by ERR_PNPM_ADDING_TO_ROOTMake the profile workspace target explicit
Fix rc.8 source builds when Node parses an ELF, shell, or Windows pnpm entrypointCross-platform pnpm entrypoint runbook
Stop black console windows flashing during Windows tool callsTwo-path Windows process-creation runbook
Find the corrupt package.json behind a pathless boot SyntaxErrorPackage-closure recovery runbook
Fix an OpenAI-compatible server_error that never retriespi-ai classification and retry evidence
Recover an old Session that throws received an update before its start MatchConversation projection recovery
Recover when pnpm leaves a plugin package behind but DSH skips reconciliationPartial plugin install recovery
Recover when a Node 24 + tsx build exits 0 but creates no artifactsSilent source-build recovery
Fix unknown or malformed prompt variables such as {{hexagon}} or {{dotted.state.path}} in tools:sdkCode Mode literal-section boundary
Fix npm ETARGET for dsh-agent-loop@^0.1.0-rc.8Registry and cache recovery
Session title stays on the first-prompt fallback with a reasoning modelAuxiliary title budget diagnosis
Generate exact installation evidence commands for my OS and install pathInteractive Install Doctor
Recover an empty sidebar or unreadable Session without destroying evidenceRoute frame, sequence, and tool-call identity failures
Stop a repeating Agent turn before it exhausts a budgetRunaway Agent loop emergency runbook
Detect repeated model text inside one streaming attemptDegenerate-output guard guide
Stop a foreground tool after session.cancel is accepted but work remains liveStuck-tool cancellation runbook
Recover when every turn in one Session returns invalid JSONPoisoned Session recovery guide
Recover when every retry reports insufficient tool messagesMissing tool-result recovery guide
Fix a todo that stays in progress after the final answerTodo state and projection guide
Fix a Code Mode Skill that appears in UI but not model contextCode Mode Skill context guide
Fix terminal inspection is unsupported on platform win32 in MinimalWindows Minimal preset Bash guide
Diagnose a frozen first workspace-write call on WindowsWindows first ACL grant guide
Run DeepSeek Harness from source on a Synology NASSynology NAS deployment guide
Recover when the Agent waits but no question or approval card appearsMissing interaction card guide
Diagnose Output token limit reached without confusing it with context overflowOutput token ceiling guide
Recover when a Git plugin installs without its declared dist/ or lib/ exportMissing plugin artifact guide
Fix /compact when its summary reaches the token capCompaction summary truncation guide
Decide which plugin missing peer and ignored-build warnings require actionPlugin peer-warning guide
Keep answers in Chinese or another language and diagnose an English Think rowResponse and reasoning language guide
Recover a long prompt after Web clears it but the turn failsAccepted-before-durable prompt recovery
Fix Bailian/DashScope UNKNOWN_TOOL or tool "" is disabled after a streamed call loses its ID and nameStreamed tool-call identity guide
Explain slow first-token latency before blaming the providerMature Session TTFT guide
Fix ReplaceFileW EACCES while editing a Windows profileWindows HMR-watched config recovery
Decide whether worker-thread Code Mode fits the security boundaryCode Mode trust-boundary guide
Fix Web, headless, or custom-profile startup reporting --expose-internals is requiredHMR loader-capability diagnosis
Fix a pnpm global install that reports an installed plugin as missingGlobal native-binding resolution guide
Fix macOS workspace selection when the path ends in :/macOS native picker path guide
Persist downstream plugin events without breaking Session resumeCustom Session event compatibility
Connect external MCP toolsMCP integration guide
Add reusable Agent instructionsSkills guide
Delegate work to child AgentsSubagents guide
Understand the runtimeThe agent-runtime mental model
Understand one complete turnAgent Loop and Session Events
Choose between Session persistence and long-term memorySessions are not long-term memory
Understand approval, guards, and tool effectsTool execution pipeline
Build an Agent rather than a loose collection of toolsAgent design map
Research a repository without publishing changesRepository Research Agent recipe
Run or debug DeepSeek Harness on WindowsWindows compatibility guide
Recover a profile after a plugin changePlugin install and recovery guide
Diagnose an ERR_HTTP2_INVALID_SESSION crashHTTP/2 provider-transport troubleshooting
Fix crypto.randomUUID or host.pickDirectory HTTP 403 without weakening the Agent shellRemote Web control-plane guide
Fix persistent Bash on NixOS or minimal LinuxPTY shell-path guide
Protect or recover a session logLive session log durability
Fix first Session flush when the filesystem rejects hard linksSession hard-link compatibility runbook
Fix a failing installation or runTroubleshooting index
Track upstream changesUpdates and breaking changes

The agent-first mental model

flowchart LR
  U[User goal] --> A[Agent contract]
  A --> C[Profile + Bundles + Patches]
  C --> G[Cordis plugin graph]
  G --> L[Agent Loop]
  L --> M[Model provider]
  L --> T[Tools + policy + approval + sandbox]
  L --> S[Durable Session events]
  S --> L
  S --> H[Web, headless, SDK, clients]

An agent is not just a prompt. A useful Agent has a task boundary, allowed effects, completion condition, model route, tool surface, permission policy, session strategy, failure behavior, and an operator-visible result. DeepSeek Harness supplies the runtime vocabulary for assembling those responsibilities without forcing every product into one fixed loop or interface.

What makes this handbook different

  • Agent-first: concepts are organized around building, running, and debugging Agents.
  • Source-backed: version-sensitive claims link to official documentation or source.
  • Operational: every tutorial includes success evidence, failure branches, and safety boundaries.
  • Visual: architecture pages prioritize diagrams over walls of text.
  • Living: updates, breaking changes, and troubleshooting pages follow upstream development.
  • Multilingual by design: English is canonical; translations declare their source revision and review status. Current depth is reported explicitly.

Language coverage

LocaleCurrent statusPublished coverage
EnglishCanonical115 pages
简体中文ReviewedNavigation plus three core guides
日本語DraftNavigation only
한국어DraftNavigation only
EspañolDraftNavigation only

The locale links at the top do not imply feature parity. English remains the source of truth until a translation points to the current canonical revision and has been reviewed by a fluent contributor.

Published guide map

Every item below is available now. Planned coverage lives in the public roadmap.

Getting started

Architecture

Agent patterns

Recipes

Operations

Official examples

  • Choose the right upstream example
  • Headless CLI task runner
  • Python SDK and JSON-RPC runtime
  • ACP automation server
  • MCP memory overlays
  • Self-modifying Cordis composition
  • Session-local schedules

Integrations

Plugin development

Security

Searchable operations

Repository structure

docs/<locale>/
  getting-started/     installation and first runs
  architecture/        runtime and lifecycle explanations
  agent-patterns/      design decisions for real agents
  recipes/             reproducible agent builds
  troubleshooting/     symptom-driven diagnostic pages
  ecosystem/           plugins, tools, skills, and comparisons
  updates/             upstream change coverage
scripts/               content and translation verification
content-manifest.json  canonical revision and locale status

Editorial and commercial boundary

DeepSeek Harness remains the subject of every technical page. SandBase maintains the handbook and may provide a restrained link to related Agent, model, Skill, or MCP discovery resources. A mention is never presented as an official DeepSeek recommendation, a compatibility guarantee, or a security endorsement.

Contributing

Corrections, reproducible examples, diagrams, troubleshooting cases, upstream change notes, and fluent translation reviews are welcome. Read CONTRIBUTING.md and run npm run check before submitting a pull request.

New here? Choose a scoped task from the public roadmap, or open a documentation request. Reproducible evidence is more valuable than a large patch.

Primary sources

License

Apache-2.0. See LICENSE.