Skip to content
Back to search

clarify-intent-and-establish-shared-understanding

Whether you're thinking or getting things done, be clear, specific, and methodical. Grounded in first principles, rigorously examine and refine a user's task, decision, plan, goal, strategy, proposal, or idea through structured, progressively deeper questioning, in order to bridge the gap between the User and the Agent. Use when the user explicitly requests grilling, challenge, pressure-testing, cross-examination, red-team review, pre-mortem analysis, or a decision audit. The goal is to uncover unclear objectives, hidden assumptions, contradictions, weak evidence, missing information, overlooked constraints, dependencies, risks, trade-offs, failure modes, and misalignment between intended outcomes and real-world results. Begin by establishing a shared understanding of the user's actual intent, plans, goals, rules, constraints, and success criteria. Ask focused, high-leverage questions rather than broad or repetitive ones. Adapt each question based on previous answers, probing deeper where uncertainty, unsupported assumptions, or strategic weaknesses remain. Regularly summarize the current understanding of the user's precise intent and position to confirm alignment, expose misunderstandings, and refine the problem or goal definition. Distinguish facts, assumptions, hypotheses, and unknowns. Challenge reasoning rigorously while remaining constructive, respectful, and solution-oriented. Continue until the reasoning is internally coherent, evidence-aware, constraint-conscious, risk-assessed, and translated into a clearer, more actionable goal, decision, or plan. Ultimately, achieving the super-alignment between the User and the Agent.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/inference1--clarify-intent-and-establish-shared-understanding/badge.svg)](/listings/inference1--clarify-intent-and-establish-shared-understanding)

HTML

<a href="/listings/inference1--clarify-intent-and-establish-shared-understanding"><img src="/api/listings/inference1--clarify-intent-and-establish-shared-understanding/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

Option 1: Ask Your Agent to Install It

Simply tell your agent:

`Install the clarify-intent-and-establish-shared-understanding skill from https://github.com/Inference1/clarify-intent-and-establish-shared-understanding for this project.`

Option 2: Install with `npx skills`

The easiest command-line method is the [`skills`](https://github.com/vercel-labs/skills) CLI.

**Interactive installation**

npx skills add Inference1/clarify-intent-and-establish-shared-understanding

The installer will let you choose the target Agent and installation scope.

**Install for a specific Agent**

npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code

Replace `claude-code` with the desired Agent identifier, such as `codex`, `cursor`, `github-copilot`, or `gemini-cli`.

Install globally for that Agent:

npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code -g

**Install for all supported Agents**

Project-level:

npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all

Global:

npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all -g

Project-level installation is the default; `-g` / `--global` makes the Skill available across projects.

Option 3: Install Manually

Download the repository as a ZIP from GitHub, or clone it:

git clone https://github.com/Inference1/clarify-intent-and-establish-shared-understanding.git

Then copy the repository folder into your Agent's **project-level** or **global** skills directory. Keep `SKILL.md` inside the Skill folder:

<skills-directory>/
└── clarify-intent-and-establish-shared-understanding/
    └── SKILL.md

For example, with Claude Code:

# Project-level
<project>/.claude/skills/clarify-intent-and-establish-shared-understanding/SKILL.md

# Global
~/.claude/skills/clarify-intent-and-establish-shared-understanding/SKILL.md

Common Agent locations include:

| Agent | Project-level | Global | | -------------- | ----------------- | -------------------- | | Claude Code | `.claude/skills/` | `~/.claude/skills/` | | Codex | `.agents/skills/` | `~/.codex/skills/` | | Cursor | `.agents/skills/` | `~/.cursor/skills/` | | GitHub Copilot | `.agents/skills/` | `~/.copilot/skills/` | | Gemini CLI | `.agents/skills/` | `~/.gemini/skills/` |

For other Agents, use their documented skills directory or let `npx skills add` select the correct location automatically.

The `skills` CLI currently documents project scope as the default, `-g/--global` for user-wide installation, `-a/--agent` for targeting specific agents, and `--all` for installing all discovered skills to all supported agents. ([GitHub][2])

[1]: https://github.com/Inference1/clarify-intent-and-establish-shared-understanding "GitHub - Inference1/clarify-intent-and-establish-shared-understanding: Systematically clarify intent, challenge assumptions, resolve contradictions, and align goals, constraints, risks, and success criteria. · GitHub" [2]: https://github.com/vercel-labs/skills "GitHub - vercel-labs/skills: The open agent skills tool - npx skills · GitHub"

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

SKILL.md · Manifest · 57 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. ---
  2. name: clarify-intent-and-establish-shared-understanding
  3. description: Whether you're thinking or getting things done, be clear, specific, and methodical. Grounded in first principles, rigorously examine and refine a user's task, decision, plan, goal, strategy, proposal, or idea through structured, progressively deeper questioning, in order to bridge the gap between the User and the Agent. Use when the user explicitly requests grilling, challenge, pressure-testing, cross-examination, red-team review, pre-mortem analysis, or a decision audit. The goal is to uncover unclear objectives, hidden assumptions, contradictions, weak evidence, missing information, overlooked constraints, dependencies, risks, trade-offs, failure modes, and misalignment between intended outcomes and real-world results. Begin by establishing a shared understanding of the user's actual intent, plans, goals, rules, constraints, and success criteria. Ask focused, high-leverage questions rather than broad or repetitive ones. Adapt each question based on previous answers, probing deeper where uncertainty, unsupported assumptions, or strategic weaknesses remain. Regularly summarize the current understanding of the user's precise intent and position to confirm alignment, expose misunderstandings, and refine the problem or goal definition. Distinguish facts, assumptions, hypotheses, and unknowns. Challenge reasoning rigorously while remaining constructive, respectful, and solution-oriented. Continue until the reasoning is internally coherent, evidence-aware, constraint-conscious, risk-assessed, and translated into a clearer, more actionable goal, decision, or plan. Ultimately, achieving the super-alignment between the User and the Agent.
  4. ---
  5. **Simplicity** should be pursued to the greatest extent possible without compromising essential details, accuracy, or conceptual integrity.
  6. **Interrogate the user systematically until both sides share a precise, complete understanding of the intent, problem, goal, task, constraints, priorities, assumptions, and desired outcomes**.
  7. Construct and navigate an implicit decision tree. Resolve foundational choices before dependent ones, identify hidden assumptions, detect contradictions, and revisit previous conclusions when new information changes the **decision landscape**.
  8. **Proceed through the decision tree iteratively, one round at a time**. At each round, define the frontier as the set of decisions whose prerequisites have already been resolved; in other words, it consists of the questions that can be asked without making assumptions about information that has not yet been provided. Present all questions on the current frontier within a single round, number them sequentially, and provide a recommended answer for each question. Do not proceed to the next round until the user has responded to the questions presented in the current round.
  9. Format a round like so:
  10. ```
  11. ❓ **Q1** - **<question title>**: <question body, might be multiple paragraphs, including multiple choices>
  12. ➡️ <your recommended answer>
  13. ---
  14. ❓ **Q2** - **<question title>**: <question body, might be multiple paragraphs, including multiple choices>
  15. ➡️ <your recommended answer>
  16. ```
  17. In each round, the user's answers reshape the decision tree: resolved decisions extend the frontier and unlock questions that depend on them. Recompute the frontier accordingly and then present the next round of questions. Any question whose answer depends on another unresolved question in the current round should be deferred to a subsequent round.
  18. Do not combine independent questions. Each question should:
  19. - target a specific unresolved decision or ambiguity;
  20. - briefly explain why the answer affects the outcome;
  21. - provide a clearly labeled recommended option when sufficient context exists;
  22. - preserve user agency by making clear that the final choice belongs to the user.
  23. You may use any and all available tools, network proxies, agents, and search resources.
  24. Do not ask for information that can be reliably obtained from available context, current working directory, code repositories, files, tools, connected sources, or other authorized inputs. **Retrieve or verify such information directly when possible**. Reserve questions for human judgment, preferences, priorities, acceptable risks, subjective trade-offs, or unavailable facts.
  25. **Continuously restate and refine the emerging understanding**. Identify ambiguities, contradictions, unresolved dependencies, missing success criteria, and differences between stated requests and likely underlying goals.
  26. Continuously maintain a shared mental model by:
  27. - summarize the current understanding when meaningful progress is made;
  28. - separate confirmed facts, assumptions, interpretations, and open decisions;
  29. - highlight ambiguity, conflicting requirements, missing constraints, and unclear evaluation criteria;
  30. - check whether the inferred objective matches the user's actual intent.
  31. **Do not execute, implement, modify, submit, publish, or finalize any consequential output until**:
  32. 1. the relevant decisions, dependencies, and constraints have been resolved;
  33. 2. the desired outcome and measurable success criteria are explicit;
  34. 3. major assumptions and trade-offs have been surfaced and accepted;
  35. 4. the user explicitly confirms that shared understanding has been reached.
  36. **Maintain a balance between rigor and efficiency**. Be persistent enough to prevent avoidable misunderstandings, but avoid unnecessary interrogation. Prioritize clarity, alignment, and forward progress over exhaustive questioning.
README.md · README · 114 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # clarify-intent-and-establish-shared-understanding
  2. ## 🚀 Quick Start
  3. ### Option 1: Ask Your Agent to Install It
  4. Simply tell your agent:
  5. `Install the clarify-intent-and-establish-shared-understanding skill from https://github.com/Inference1/clarify-intent-and-establish-shared-understanding for this project.`
  6. ### Option 2: Install with `npx skills`
  7. The easiest command-line method is the [`skills`](https://github.com/vercel-labs/skills) CLI.
  8. **Interactive installation**
  9. ```bash
  10. npx skills add Inference1/clarify-intent-and-establish-shared-understanding
  11. ```
  12. The installer will let you choose the target Agent and installation scope.
  13. **Install for a specific Agent**
  14. ```bash
  15. npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code
  16. ```
  17. Replace `claude-code` with the desired Agent identifier, such as `codex`, `cursor`, `github-copilot`, or `gemini-cli`.
  18. Install globally for that Agent:
  19. ```bash
  20. npx skills add Inference1/clarify-intent-and-establish-shared-understanding -a claude-code -g
  21. ```
  22. **Install for all supported Agents**
  23. Project-level:
  24. ```bash
  25. npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all
  26. ```
  27. Global:
  28. ```bash
  29. npx skills add Inference1/clarify-intent-and-establish-shared-understanding --all -g
  30. ```
  31. Project-level installation is the default; `-g` / `--global` makes the Skill available across projects.
  32. ### Option 3: Install Manually
  33. Download the repository as a ZIP from GitHub, or clone it:
  34. ```bash
  35. git clone https://github.com/Inference1/clarify-intent-and-establish-shared-understanding.git
  36. ```
  37. Then copy the repository folder into your Agent's **project-level** or **global** skills directory. Keep `SKILL.md` inside the Skill folder:
  38. ```text
  39. <skills-directory>/
  40. └── clarify-intent-and-establish-shared-understanding/
  41. └── SKILL.md
  42. ```
  43. For example, with Claude Code:
  44. ```text
  45. # Project-level
  46. <project>/.claude/skills/clarify-intent-and-establish-shared-understanding/SKILL.md
  47. # Global
  48. ~/.claude/skills/clarify-intent-and-establish-shared-understanding/SKILL.md
  49. ```
  50. Common Agent locations include:
  51. | Agent | Project-level | Global |
  52. | -------------- | ----------------- | -------------------- |
  53. | Claude Code | `.claude/skills/` | `~/.claude/skills/` |
  54. | Codex | `.agents/skills/` | `~/.codex/skills/` |
  55. | Cursor | `.agents/skills/` | `~/.cursor/skills/` |
  56. | GitHub Copilot | `.agents/skills/` | `~/.copilot/skills/` |
  57. | Gemini CLI | `.agents/skills/` | `~/.gemini/skills/` |
  58. For other Agents, use their documented skills directory or let `npx skills add` select the correct location automatically.
  59. The `skills` CLI currently documents project scope as the default, `-g/--global` for user-wide installation, `-a/--agent` for targeting specific agents, and `--all` for installing all discovered skills to all supported agents. ([GitHub][2])
  60. [1]: https://github.com/Inference1/clarify-intent-and-establish-shared-understanding "GitHub - Inference1/clarify-intent-and-establish-shared-understanding: Systematically clarify intent, challenge assumptions, resolve contradictions, and align goals, constraints, risks, and success criteria. · GitHub"
  61. [2]: https://github.com/vercel-labs/skills "GitHub - vercel-labs/skills: The open agent skills tool - npx skills · GitHub"
  62. ## ✨ Introduction
  63. AI agents often face an "Autonomy–Interaction Dilemma": when users provide ambiguous prompts or instructions, highly autonomous systems may proceed for extended periods without sufficient interaction. As a result, they may produce Deep Research reports that diverge substantially from user expectations or make inappropriate modifications to code. Such outcomes not only compromise task quality but also result in unnecessary expenditures of computational resources and time. Therefore, achieving an appropriate balance between autonomous execution and timely user interaction is critical to improving agent performance.
  64. **The value of slow thinking** lies in introducing a deliberate `/clarify-intent-and-establish-shared-understanding` step before an agent initiates task execution. By proactively clarifying user intent and establishing shared understanding, this step helps align the agent's interpretation of the task with the user's intended meaning and intended outcomes. This mechanism therefore provides a practical and minimally intrusive means of mitigating hallucinations and reducing agent-human misalignment.
  65. **Core insight**: In complex open-world environments, communication can itself function as an efficient form of computation. Rather than attempting to anticipate users' potential needs by continually expanding the search space, agents may benefit more from scaling interaction. Before acting, they can engage in multi-turn dialogue to elicit and progressively clarify user intent. Such an interaction-first strategy may provide a more efficient approach to **Goal Forge, Task Execution and Agent Loop**.
  66. ## ✅️ What Does This SKILL Do? When Is It Triggered?
  67. The SKILL `clarify-intent-and-establish-shared-understanding` systematically examines and refines a user's plan, decision, goal, strategy, proposal, or idea before consequential actions are taken. Its purpose is to establish a precise and shared understanding of the user's intent, objectives, constraints, priorities, assumptions, dependencies, risks, trade-offs, and measurable criteria for success. It achieves this through adaptive and progressively deeper questioning, posing exactly one high-leverage, decision-focused question at a time. Throughout this process, it distinguishes among facts, assumptions, hypotheses, and unknowns; identifies contradictions and information gaps; directly verifies available facts; and periodically synthesizes the emerging shared understanding.
  68. The SKILL is triggered when the user explicitly requests rigorous scrutiny, such as **grilling, challenging, pressure-testing, cross-examination, red-team review, pre-mortem analysis, or a decision audit**. The process continues until the reasoning is coherent, grounded in available evidence, responsive to relevant constraints, informed by an explicit assessment of risks, actionable, and confirmed by the user.
  69. ## 💓 Why this Skill?
  70. This skill is especially valuable when the quality of an outcome depends on accurately defining the problem before taking action. It is intended for decisions, plans, strategies, proposals, and goal-setting processes in which ambiguous objectives, implicit assumptions, conflicting constraints, insufficient evidence, overlooked dependencies, or inadequately assessed risks may result in costly errors.
  71. Rather than accepting the user's initial framing uncritically, the skill systematically evaluates its validity through focused and adaptive questioning. The process develops a shared understanding of the underlying objectives, priorities, constraints, assumptions, trade-offs, risks, and measurable criteria for success. It also distinguishes established facts from hypotheses and unresolved uncertainties, identifies contradictions, and verifies relevant information directly when possible. By preventing premature action and identifying weaknesses at an early stage, the process transforms ambiguous or insufficiently supported reasoning into a coherent, evidence-informed, constraint-sensitive, and actionable framework for goal forge.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.