ipollowork

DeepSeek Harness 可编辑设计系统：AI 生成、可视化编辑、模板市场与 PPT｜Native Design & PPT Studio for DeepSeek Harness.

Jev DSH 决策引擎｜面向 Agent Harness 的结构化决策插件。原生支持 DeepSeek Harness，通过 iPolloWork 支持 OpenCode、Codex Harness。

iPollo Onto · 业务本体 — 轻量、AI 原生的业务数据底座。独立运行，连接表格、关系与文档；作为 iPolloWork 插件使用。

Offline OpenTelemetry trace triage for AI-agent errors, latency, tool activity and broken parent links, packaged for iPolloWork.

Audit synthetic agent-training tasks for environment feasibility, verifier quality, evidence gaps, and batch coverage.

Community directory for iPolloWork plugins discovered through the ipollowork-plugin topic. · iPolloWork 插件社区目录

Evaluate Claude Sonnet model migrations with paired workload evidence, critical quality gates, normalized token costs, compatibility checks, and reversible rollout plans.

Audit GPT-6.1 Sol API requests for endpoint, tool, reasoning, residency, context and token-cost compatibility before migration.

Audit AI-assisted web research with claim-level source mapping, official task links, freshness checks, and prioritized verification routes.

Least-privilege sandbox reviews for local coding agents: map filesystem, network, credentials and session lifecycle without inventing unsupported controls.

AI review evidence audit Skill and iPolloWork plugin — verify Copilot findings against code and tests. 中英双语 AI 审查复核技能，支持 iPolloWork。

Plan AI research validation within a time budget: predictions, falsifiers, observations and JSON/Markdown handoffs. 支持 iPolloWork 的中英双语假设验证队列插件。

Evidence-led code-review handoffs for GitHub Copilot and agent-assisted changes, packaged as an iPolloWork Skill.

Evidence-led model rollout cards for GitHub Copilot, packaged as an iPolloWork Skill.

Evidence-bound reviews for GitHub Copilot Memory: classify repository facts, flag stale or sensitive entries, and map risks across agentic autofix and Copilot agents.

Offline AI review evidence board for iPolloWork: import JSON, filter findings, export Markdown. 中英双语审查证据看板插件。

Turn AI-assisted fuzzing crash artifacts into evidence-preserving duplicate clusters, reproduction states, and validation handoffs.

Review agent-generated security-fix pull requests with alert traceability, CodeQL and behavioral evidence, scope-risk gates, and human-owned merge decisions.

Evidence-based test planning for Claude Code and AI-assisted changes: audit map freshness, preserve required checks, export JSON/Markdown. 支持 iPolloWork 的测试影响证据 Skill。

Evidence-linked review of AI-assisted code migrations: contracts, lifecycle traces and scoped approvals. iPolloWork Skill · 中英双语迁移行为核对。