DeepSeek Harness Plugins
DeepSeek Harness PluginManifest Valid

dsh-plugin-abtest

Paired A/B testing framework and deterministic promotion gates for DeepSeek Harness plugins.

Developer ToolsSecurity & PolicySkills & WorkflowsRemote Execution
0GitHub Stars0ForksUpdated2026-08-26

Install

$ dsh plugin --profile web add 'github:Morriaty-The-Murderer/dsh-plugin-abtest#<full-commit-sha>'

Plugin Overview & Capabilities

AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.

source-grounded
dsh-plugin-abtest provides reproducible paired A/B experiment evaluation for DeepSeek Harness plugins. It runs control and candidate variants across isolated environments, pairs results case by case, and produces deterministic promotion recommendations (PROMOTE, REVIEW, REJECT, INCONCLUSIVE) based on auditable evidence and configured metric thresholds.

Key Capabilities

  • Execute paired benchmark evaluations between control and candidate plugin versions in isolated DSH runners
  • Evaluate task success, token consumption, P95 latency, and tool error rates with blind comparisons
  • Generate deterministic promotion decisions (PROMOTE, REVIEW, REJECT, INCONCLUSIVE) using manifest-defined rules
  • Export comprehensive auditable evidence packs including JSON, Markdown, and static HTML reports
  • Integrate into DSH profiles via Cordis extension manifest or execute standalone via CLI

Useful For

  • Validating prompt, tool, or code changes in a DSH plugin before shipping to production
  • Benchmarking latency and token efficiency regressions across plugin updates
  • Implementing CI/CD promotion gates for LLM agent plugins with auditable offline evidence

Who It Fits

  • DeepSeek Harness plugin developers
  • Agent evaluation and QA engineers
  • LLM application platform maintainers

Documented Limitations

  • Does not automatically mutate production profiles or publish plugins; all promotion recommendations are offline only
  • Requires Node.js ^22.19.0 or >=24.0.0 and DSH 0.1.0-rc.7 contract compatibility
  • Model-facing tool entry points disallow test commands containing arbitrary command executions

DSH Compatibility

Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.

Not tested yetNo runtime compatibility tests have been published yet.

Security Signals

Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.

Dependency Counts

package.json declares 4 runtime, 9 development, 2 peer, and 0 optional dependencies.

info
Dsh Bundle Declared

package.json declares DSH bundle metadata.

info
License Declared

GitHub reports the repository license as MIT.

info
Package Manifest Available

A root package.json was captured and can be inspected by the registry.

info
Source Available

Public GitHub source metadata is available for this registry snapshot.

info

Source & Registry Notes

Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.

Source repository
Morriaty-The-Murderer/dsh-plugin-abtest
Registry source
GitHub · dsh-plugin topic
Registry classification
Plugin
Source checked
bfff76f · 2026-08-28

This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.

Repository Activity

GitHub stars
0GitHub stars
Forks
0
Open issues
0
Last commit
2026-08-26
Last release
2026-08-26
For maintainers

Maintaining this plugin?

This listing is generated from public repository data. If you maintain this project, you can review the information and share this listing with your users if you find it useful.

Add to README
Listed on DSHPlugin.app