Apache-2.0 agent skill for DeepSeek Harness and other coding agents: a pre-execution clarification step that aligns intent, audits assumptions and contradictions, and confirms measurable success criteria before any action.
DSH integration
Compatible
Author-claimed
Safety audit
Unaudited
Last verified
2026-08-23
License
Apache-2.0
01What can it help you accomplish?
Add a deliberate /clarify-intent-and-establish-shared-understanding step before an agent starts executing, so ambiguous prompts get resolved first
A shared understanding of objectives, priorities, constraints, assumptions, trade-offs, risks and measurable success criteria — confirmed by the user before any action
Users and agents who want to reduce hallucinations and human-agent misalignment on decisions, plans, strategies, proposals and goal-setting, especially when prompts are ambiguous
02How to install into DeepSeek Harness
Prerequisites
- A coding agent that supports a skills directory (Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, DeepSeek Harness, etc.), or Node.js with the `npx skills` CLI
Installation steps
- 01
Easiest: `npx skills add Inference1/clarify-intent-and-establish-shared-understanding` (interactive; picks target Agent and scope)
$ npx skills add Inference1/clarify-intent-and-establish-shared-understanding
- 02
For a specific agent: `npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code` (replace with codex, cursor, github-copilot, gemini-cli, etc.); add -g for global
$ npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code
- 03
For all supported agents: `npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all` (add -g for global)
$ npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all
- 04
Manual: clone the repo and copy the folder (keeping SKILL.md) into the agent's project-level or global skills directory
Verify the integration
Not specified by the author
03DSH integration and capability boundaries
Compatible agent skill — installed via the `npx skills` CLI or by copying SKILL.md into an agent's skills directory; works with DeepSeek Harness and other coding agents that support the skills directory convention
Intent clarification & shared understanding
An ambiguous user prompt or task before execution→Multi-turn clarification that surfaces objectives, priorities, constraints, assumptions, trade-offs and risks; the reasoning continues until it is coherent, evidence-grounded, constraint-sensitive and user-confirmed
Assumption & contradiction audit
The user's initial framing of a decision, plan or goal→Distinguishes established facts from hypotheses and unresolved uncertainties, identifies contradictions, and verifies relevant information when possible
04Who is it for? When not to use it?
Good for
- Users and agents who want to reduce hallucinations and human-agent misalignment on decisions, plans, strategies, proposals and goal-setting, especially when prompts are ambiguous
05Compatibility, maintenance and safety notes
- This is a Skill, not a bundled DSH plugin: it is installed into an agent's skills directory and invoked as a step; it adds no runtime, port, or service of its own.
Apache-2.0 · no tagged release yet (repo pushed 2026-08-15)
06Frequently asked questions
How do I install it for DeepSeek Harness?
Use the `npx skills` CLI: `npx skills add Inference1/clarify-intent-and-establish-shared-understanding` lets the installer pick the target Agent and scope. Or clone the repo and copy the folder (keeping SKILL.md) into your agent's project-level or global skills directory.
What does the skill actually do?
Before the agent acts, it runs multi-turn clarification to surface objectives, priorities, constraints, assumptions, trade-offs and risks, distinguishes facts from hypotheses, finds contradictions, and keeps going until the reasoning is coherent, evidence-grounded and user-confirmed.
Does it add any runtime or service?
No. It is a Skill installed into an agent's skills directory and invoked as a step; it adds no runtime, port, or background service of its own.
Why bother with a clarification step?
Highly autonomous agents can run for a long time on ambiguous prompts and then produce Deep Research reports or code changes that diverge from what you meant. A short interaction-first step is a practical, minimally intrusive way to mitigate hallucinations and reduce misalignment.
07Related DSH workflows
distilly
by titanwings
Meta-skill turning source material into a versioned Person Profile skill with incremental merge, correction layer, rollback, per-host install.
dsh-market
by dsh-market
插件市场 (plugin marketplace): 打开 Settings → Plugin Market 浏览/搜索/一键安装社区插件；安装源限制为 curated awesome-dsh-plugin 注册表（其余拒绝），默认禁用构建脚本（pnpm>=10 需显式开启）。
ai_animation
by unclecheng-li
Generate standalone animated HTML diagrams, presentations, notes, protocol visualizations, and UI demos from prompts.
aegis
by ganyuanran
Make AI coding agents architecture-aware: baseline-first, evidence-verified, drift-checked, and safe across long tasks.
08Data and sources
npx skills add Inference1/clarify-intent-and-establish-shared-understanding
Install the clarify-intent-and-establish-shared-understanding skill from https://github.com/Inference1/clarify-intent-an…
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-08-23. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.