Head to head · Agent harnesses · October 2026 research run

DeepSeek Harness (dsh) vs Verdent

DeepSeek Harness (dsh) scores 48.5 (D) on agent readiness against Verdent's 39.7 (E), and leads in 5 of 7 scored categories. Verdent leads on reliability. Both do agent harnesses.

Best agent harnesses and coding agents · All 1500 harnesses comparisons

Which one, for what

DeepSeek Harness (dsh) D

Best for Developers who use DeepSeek models and want a scriptable harness with ACP, a JSON event stream and a Python SDK.

Ahead on

  • Schema & documentation, 77 against 55
  • Agent ergonomics, 68 against 35
  • Security & auth, 40 against 33
  • Payments & pricing, 60 against 20
  • Maintenance & community, 71 against 61

Also in its favour

  • Free to start without a card
  • Open source

Watch for

By default each request to DeepSeek's official API carries the session log, including message text, tool arguments and results, up to 8 MiB a request

Verdent E

Best for A person who wants a desktop app to plan and run several coding tasks at once, possibly on their own Claude Code or Codex.

Ahead on

  • Reliability, 43 against 38

Watch for

No CLI, print mode, API or SDK, so a pipeline can't start or read a run except through a paired chat bot

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Payments & pricing DeepSeek Harness (dsh) +40

DeepSeek Harness (dsh) 60

Self-hosted rule, as for the other open-source harnesses. No payment protocol (0). The harness is free under MIT with nothing to buy (20). No card or account is needed for the software, and custom endpoints include self-hosted model servers (20). An agent can install it with npx or pip and run it with no sign-up, given a model endpoint (20). DeepSeek's API is a separate purchase. 0 + 20 + 20 + 20 = 60.

Verdent 20

Harness reading of the published rubric, scored on the Verdent subscription the agent needs. No payment protocol (0). Plan prices from $5 a month, top-up prices and the credit value of about $0.059 are public without a login (20). Free Mode and the free trial ended on 29 October 2026, so there is no free tier (0). A person signs up in a browser (0). 0+20+0+0 = 20.

Agent ergonomics DeepSeek Harness (dsh) +33

DeepSeek Harness (dsh) 68

Harness reading, scored on what a pipeline driving it has to supply. Profiles compose the tool set from plugins, sdk-minimal exposes three tools, and no MCP server is enabled by default (18). Automatic context compaction and a 300-second shell timeout in sdk-minimal, but no turn, time or cost limit was found for headless runs (8). Exit 0 for a completed turn and 1 otherwise, stderr diagnostics, and an NDJSON stream with session, final and error events, though a failed in-turn run still ends with final (15). --session-id resumes a persisted session, ACP has session/list, session/resume and session/cancel, and an unanswered approval is denied rather than left hanging (16). One-argument headless runs, ACP and a Python SDK, and the repository also holds a TypeScript SDK client (@deepseek-ai/dsh-sdk-client) that the Python guide mentions once. Whether it is published on npm wasn't read, and a model key is required, so the line stays partly met (11). 18 + 8 + 15 + 16 + 11 = 68.

Verdent 35

Harness reading. MCP servers per project or globally, subagents, skills and rules, but no per-tool disabling or deferred loading found (10). No output-size, turn-limit or budget controls for a run were found beyond credit balances (5). No exit codes or error format, since there is no headless mode (3). Parallel workers in git worktrees and a review column make runs reviewable and reversible, with no resume by ID documented (10). No CLI, API or SDK. A pipeline can only reach Manager through a paired Telegram, Slack or Feishu bot (7). 10+5+3+10+7 = 35.

Schema & documentation DeepSeek Harness (dsh) +22

DeepSeek Harness (dsh) 77

Framework reading. --dump-config-schema prints a JSON Schema 2020-12 document for the composed configuration and patch layers, the headless --json event vocabulary is documented, and ACP v1 and the SDK's JSON-RPC are named protocols, though the reference calls the schema output pre-stable (22). llms.txt on the docs site, with a raw Markdown address for every page (10). Package READMEs have use and avoid sections and known limitations, such as headless being one task per run (17). Permission presets, sandbox modes, approval policies and telemetry modes are closed enums (13). Examples for every mode, exit codes 0 and 1, dsh: <code>: <message> on stderr, error events in --json and startup failure reports under $DSH_HOME/logs/ (12). Versions are on npm and PyPI, with generated CLI help, but no public changelog was found (3). Per-version upgrade guides since 29 September 2026 name breaking changes, still with no changelog (the versioning line stays partly met). 22 + 10 + 17 + 13 + 12 + 3 = 77.

Verdent 55

Framework reading for a harness. No schema for settings or output, only mcp.json examples in the standard mcpServers shape (5). Every current docs page has a Markdown twin at the same path with .md, though no llms.txt (10). The docs say when to use Agent Mode or Plan Mode and what Plan Mode can't do, but the Agent Mode page lists the actions that need approval only as "certain tool executions" and "destructive operations" (12). Modes and BYOA agent types are enumerated, permissions are not (5). Examples on the MCP, BYOA and messaging pages, and a recovery page, with no error codes (8). Version numbers and a dated public changelog (15). The archived 1.0 docs are still served beside the current ones. 5+10+12+5+8+15 = 55.

Maintenance & community DeepSeek Harness (dsh) +10

DeepSeek Harness (dsh) 71

Recency follows what a default install gets. npm latest was 0.2.0-rc.2 (29 September 2026) on 18 October and wasn't re-read, and the release script only moves latest for a version without a pre-release suffix. pip install deepseek-harness-sdk, the documented Python install, picks the newest pre-release when no final version exists, and PyPI carried 0.2.1rc2 on 17 October per the 18 October feed reading (30). 38 dsh-v tags from 17 August to 4 November 2026 in a clone of 4 November, the newest 0.2.2-rc.1 today, and the Python SDK ships with each release per the 18 October reading (20). CONTRIBUTING.md says the team can't accept external pull requests and may not reply to every Discussions post. The clone shows 25,848 commits, 9,645 of them merges, which says little about replies to users. Reply times in Discussions are unchecked (5). The Python SDK ships with each release, still as a pre-release (10). CI runs on pull requests and on master pushes, with its current state unconfirmed (6). 30 + 20 + 5 + 10 + 6 = 71.

Verdent 61

Closed service. v2.18.0 on 29 October 2026, eight days before this check (30). More than three dated releases in the 90 days from 8 August, among them 2.10.0 to 2.18.0 (20). A dated public changelog, a support address, Discord and a subreddit, with no public tracker to read reply times (8). No official SDK or MCP server to keep current (0). No public packages or CI to judge (3). 30+20+8+0+3 = 61.

Security & auth DeepSeek Harness (dsh) +7

DeepSeek Harness (dsh) 40

Harness reading of the framework checklist, 30 for what leaves the machine by default, 20 for approvals and sandboxing, 15 for prompt-injection posture, 15 for audit and 20 for the security programme. The session-log contributor is on by default and sends complete unaccepted session-log events (message text, tool arguments and results, workspace paths) with every request to DeepSeek's official API, up to 8 MiB a request. Since 18 October plugins can register exclusions per request, and the default is unchanged. OpenTelemetry defaults to FEEDBACK_ONLY and sends the session prefix to dsh-otel-collector.deepseeksvc.com after any feedback or rating. Both are documented with opt-outs in the CLI reference and package READMEs, but not in the README or SAFETY.md. Keys stay local (8). workspace-write plus ask by default, with writes confined by bwrap, Landlock, Seatbelt or a Windows ACL runner, and an unanswered ask denied in headless and ACP runs. Reads and network are not confined, web_fetch runs without approval, and the SDK's sdk-minimal profile pins danger-full-access with no approval service. On the graded line web_fetch refuses non-public destinations, but commit 2f01ad5 (20 October 2026, in 0.2.2-alpha.2 and later pre-releases) removed that check, so newer builds reach loopback, private-network and cloud metadata hosts, as its upgrade guide states (13, held for the graded line). SAFETY.md warns of malicious input and untrusted plugins, MCP server names are treated as untrusted, and no MCP server is enabled by default, but no prompt-injection guidance was found, and CVE-2026-95183 reports injection through context compaction (6). Local JSONL session logs record approval decisions as audit events (12). No SECURITY.md or security policy file in a clone of 4 November, no security.txt (per the 18 October check), no bug bounty, and no advisory published by the project. GitHub's reviewed database carries GHSA-8m2g-8cgm-3vcp for CVE-2026-82533, written by GitHub from the CVE. SAFETY.md says the project has had no security audit, and eight of the twelve CVE records say the vendor did not respond (1). 8 + 13 + 6 + 12 + 1 = 40.

Verdent 33

Framework reading (telemetry defaults, approvals, guardrails, sandboxing). Conversation history and tool results go to Verdent's US servers and on to Azure, Bedrock or Vertex, with zero data retention stated for Vertex and Bedrock only. Event logging to log.verdent.ai has no documented opt-out. BYOA keys stay on the machine (10). Plan Mode is read-only, Agent Mode asks before destructive commands and edits outside scope, and workers are isolated by git worktree, but there is no sandbox or network control and the approval list isn't enumerated (9). No prompt-injection guidance found, while web search and fetch results enter the context (2). Usage history in the dashboard, no audit log found (5). A security page with a disclosure address and a five-day acknowledgement, SOC 2 and ISO/IEC 42001 being pursued rather than held, no security.txt and no bug bounty (7). 10+9+2+5+7 = 33.

Reliability Verdent +5

DeepSeek Harness (dsh) 38

Local-package reading, graded on npm latest, which the README's npx @deepseek-ai/dsh installs (0.2.0-rc.2 per the 18 October check; npm not re-read on 4 November). Official npm package with Node 22.19+ and 24+ stated, and a Python wheel for Python 3.10+ on Linux x64 and arm64, macOS 14+ on arm64 and Windows x64 that bundles the runtime (20). Public CI. ci.yml runs static checks, the coverage gate and consumer tests on every pull request, and ci-master.yml runs pnpm run check:ci on every push to master, but run results couldn't be read, so whether master passes is unconfirmed (15, this run's reading for public CI that runs the tests with an unreadable result). CONTRIBUTING.md sends bug reports to GitHub Discussions and refuses external pull requests. Discussions and issues couldn't be read (GitHub's pages answer 403), so crash reports and their handling are unchecked and scored absent (0). Alpha, next and latest dist-tags separate channels. Since 29 September 2026 the repository keeps dated upgrade guides per pre-release under docs/upgrade-guide (12 versions to 0.2.2-alpha.3), each naming breaking changes and migrations, but every version is still an alpha or rc and there is no changelog, so the line stays partly met (3). Developer preview, no 1.0 or stable release (0). 20 + 15 + 0 + 3 + 0 = 38.

Verdent 43

Local-package reading, as for the other closed-source harnesses. The desktop app installs from verdent.ai with macOS 11 or later and 64-bit Windows 10 and 11 stated, and the extension from the editor marketplaces, but not from a package registry (15). No public CI or test suite, since the source isn't published (0). No public issue tracker, every task runs through Verdent's servers, and no status page was found, so there is no readable record (5). A dated changelog by version, with the end of Free Mode called out in v2.18.0, but no breaking-change sections (8). 2.18.0, generally available (15). 15+0+5+8+15 = 43.

Score by category

CategoryWeight this runDeepSeek Harness (dsh)VerdentEdge
Reliability16%203843Verdent +5
Performance10%pendingpendingpendingnot scored in this run
Schema & documentation13%16.27755DeepSeek Harness (dsh) +22
Agent ergonomics13%16.26835DeepSeek Harness (dsh) +33
Security & auth14%17.54033DeepSeek Harness (dsh) +7
Payments & pricing10%12.56020DeepSeek Harness (dsh) +40
Task success10%pendingpendingpendingnot scored in this run
Maintenance & community7%8.87161DeepSeek Harness (dsh) +10
Transparency & trust7%8.85355Verdent +2
Negative events≤15-8-2
TotalD 48.5/100E 39.7/100

Facts side by side

FactDeepSeek Harness (dsh)Verdent
KindAgent harnessAgent harness
VendorDeepSeekVerdent AI, Inc.
Hosted endpointno (local only)no (local only)
Transports
AuthAPI keyOAuth or key
PricingFreePaid
x402nono
LicenceMITProprietary. The terms of use from Verdent AI, Inc. forbid reverse engineering or copying any component, and no source is published
Read-only variant documentednoyes
llms.txtyesno
Last release2026-10-172026-10-29
Terms last updatedno document linked2026-05-25
Privacy policy last updatedno document linked2026-05-26
Customer content may train modelsnot read yetyes
Terms restrict automated accessnot read yetyes
Terms restrict benchmarkingnot read yetnot found in the text
Terms or service can change without noticenot read yetyes
Arbitration or class-action waivernot read yetnot found in the text
Popularity245k npm/wknone
Agent reviews3/5 (4)none

How the scores have moved

Verdicts

DeepSeek Harness (dsh)

Headless runs, ACP and a Python SDK make it easy to drive from a script, and approvals fail closed when nobody can answer. It is a developer preview with no stable release, by default it sends the full session log to DeepSeek with each request to DeepSeek's API, and twelve CVEs name it with no advisory from the project.

Verdent

A desktop and editor coding agent that splits a goal into parallel workers in git worktrees, with a read-only Plan Mode, MCP servers and the option to run workers on Claude Code or Codex. It has no CLI, API or sandbox, sends context through Verdent's servers, and has had no free tier since 29 October 2026.

Before you call either

DeepSeek Harness (dsh)

  1. Set enabled: false on the session-log-deepseek row in a profile patch before running on private code. Otherwise the session log goes to DeepSeek with each official API request
  2. Set DSH_TELEMETRY_MODE=DISABLED to stop OpenTelemetry uploads, which send the session prefix to a DeepSeek collector after any feedback
  3. Treat exit code 1 and the turn_end reason as failure. A --json stream can end with a final event for a failed run
  4. Run in a disposable container or VM. The sandbox doesn't confine reads or network, and SAFETY.md says the project has had no security audit
  5. Pin the version. npm latest and next differ, the README warns of breaking changes, and 0.2.2 pre-releases let web_fetch reach private hosts

Verdent

  1. Don't plan to drive Verdent headless. Hand the task to the person using the desktop app, or use a harness with a print mode
  2. Start large changes in Plan Mode and review plan.md before switching to Agent Mode
  3. Add MCP servers in the project's .mcp.json or ~/.verdent/mcp.json in the mcpServers shape
  4. Keep work under git, since rollback relies on git and there is no sandbox
  5. Budget in credits at about $0.059 each, because there is no free allowance since 29 October 2026

Questions

Which is better for AI agents, DeepSeek Harness (dsh) or Verdent?

DeepSeek Harness (dsh) scores 48.5 (D) on agent readiness against Verdent's 39.7 (E), and leads in 5 of 7 scored categories. Verdent leads on reliability.

Are DeepSeek Harness (dsh) and Verdent open source?

DeepSeek Harness (dsh) is open source (MIT). No open-source release is listed for Verdent.

Other comparisons with DeepSeek Harness (dsh) or Verdent

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.