clarify-goal
Skill9 starsChanged 2 months agoApache-2.0View on GitHub
This skill asks questions to define a research goal. You use it when the scope, assumptions, or success criteria are unclear.
In this repository10 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- analyze-paperSkill
- brainstormSkill
- critiqueSkill
- formalize-problemSkill
- investigateSkill
- reaperSkill
- review-literatureSkill
- search-paperSkill
- write-paperSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 3 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈3
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 28
If the user supplies a paper, you scan it for this information:
- Very specific: line 53
After the user responds, you combine their answers with defaults for unanswered questions.
- Distinctive: line 72
If the PDF is unreadable, you report the error.
- Distinctive: line 76
If the goal is unclear, you ask questions and request confirmation.
- Very specific: line 77
If the goal is precise, you can omit questions and confirmation.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Clarify Goal
You must read and apply the language rules before you write any output.
Usage
You invoke this skill by name with a quoted goal and an optional paper path. Slash-command hosts use /clarify-goal "<goal>".
clarify-goal "explore the feasibility of post-quantum threshold signatures"
clarify-goal "is this protocol secure?" path/to/paper.pdfThe quoted research goal is required. An additional path to an existing file identifies the paper.
Instructions
1. Read the Paper
If the user supplies a paper, you scan it for this information:
- The paper's topic, system, and protocol.
- The paper's main theorems, security properties, and performance results.
- The sections that contain each result.
You leave full analysis to the analyze-paper skill. If the user supplies no paper, you use the goal and research domain.
2. Identify Unclear Requirements
You check the goal against the paper or research domain.
- Scope: You determine which papers, sections, theorems, or protocols the goal covers.
- Definitions: You identify terms with multiple meanings, such as security under different adversary models.
- Success criteria: You determine whether the user needs a counterexample, proof gap, or performance bound.
- Assumptions: You identify assumptions that the user can change, such as synchrony or adversary type.
- Comparison: You identify the metric and baseline for each comparison.
- Target venue: You identify the conference or journal for the paper. The default is the Tier 1 venue in
../reaper/references/venue-tiers.mdthat fits the topic best.
3. Ask Questions
You ask 3-5 questions that address the paper and goal. Each question offers concrete options where possible. Each question has a default answer. You use a numbered list. You state the defaults for questions that the user does not answer.
4. Use the Answers
After the user responds, you combine their answers with defaults for unanswered questions. You write reaper-workspace/notes/clarified-goal.md with title # Clarified Research Goal and these level-two sections:
| Section | Content |
| Original Goal | You copy the user's goal exactly. |
| Clarifying Q&A | You record each question and exact answer. You label unanswered items (default) and state the default. |
| Refined Goal | You state the precise goal with the user's answers. |
| Scope | You identify the parts of the paper or topic that the research covers. |
| Key Assumptions | You state investigation assumptions. |
| Success Criteria | You state the evidence that answers the goal. |
| Target Venue | You name the venue and the deadline if the user gives one. |
5. Report
You state the refined goal. You ask the user to confirm it before the pipeline continues.
Quality Criteria
- The refined goal gives
formalize-problemenough information to proceed. - If the PDF is unreadable, you report the error. You do not invent context.
Exceptions
- If the goal is unclear, you ask questions and request confirmation. Without a paper, you ask about the domain, terms, and expected output.
- If the goal is precise, you can omit questions and confirmation. You write
clarified-goal.md, state that the goal is clear, and continue.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.