Requirements
- Target platform
- OpenClaw
- Install method
- Manual import
- Extraction
- Extract archive
- Prerequisites
- OpenClaw
- Primary doc
- SKILL.md
Structured interview to discover personal facts and generate reusable, approved statements for proposals and cover letters. Creates personalized content for...
Structured interview to discover personal facts and generate reusable, approved statements for proposals and cover letters. Creates personalized content for...
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete.
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run.
This is NOT a CV builder. This skill creates high-quality, reusable statements for proposals and cover letters by running a structured interview to extract personal-but-professional facts.
Generic cover letter and proposal writers are bland because they don't know YOU. They can't capture: Your living in Dubai and Seoul, and what that taught you about cross-cultural collaboration The robotics project you tinkered with at home that shows your passion The class that changed how you think about systems The boring job that taught you what quality really means Your unfair advantage that clients actually hire you for This skill solves that through interview → facts → statements → approval → assembly.
The skill asks structured questions to extract: Geographic and cultural context (places lived, travel, cross-cultural work) Work experience beyond the resume (lessons, standout moments, what colleagues rely on you for) Education (specific classes, formative learning, projects) Projects and proof (side projects, demos, artifacts) Awards and recognition (scholarships, competitions, certifications) Past hobbies (skills hidden in former interests) Books and influences (working philosophy) The boring stuff (what you learned from frustrating work) Proposal leverage (what you want to be hired for) Constraints (topics to avoid, boundaries)
From gathered facts, the skill drafts 3-8 candidate statements per round. Each statement: Is grounded in real facts (no invention) Is usable in proposals/cover letters Has multiple variants (Upwork short, standard cover letter, technical/proof-first) Has tags (#leadership #robotics #global-living)
You review each statement: Approve / Edit / Reject Approved statements go into statements.md Edits refine the statement and update preferences Rejections inform future drafting
When you need a proposal, the skill pulls relevant approved statements and assembles them for the target platform and company.
All data is stored INSIDE the skill folder at: skills/proposal-interview/personal/ and skills/proposal-interview/companies/
skills/proposal-interview/ personal/ <person_id_or_name>/ profile.md # Current snapshot: name, what they do, where, objectives (ALWAYS READ FIRST) user.md # Raw facts about this person (APPEND ONLY - NEVER DELETE) statements.md # Approved statements (APPEND ONLY - NEVER DELETE APPROVED STATEMENTS) preferences.md # Writing style preferences coherence.md # Conflicts, gaps, clarifications needed companies/ <company_or_initiative_slug>/ profile.md # Current snapshot: company name, what they do, where, objectives org.md # Company facts, domain, relationships (APPEND ONLY - NEVER DELETE) statements.md # Company-specific approved statements (APPEND ONLY - NEVER DELETE) preferences.md # Company-specific style preferences projects.md # (Optional) Project history with this org coherence.md # (Optional) Company-specific conflicts/gaps
Personal vs Company: Personal facts → skills/proposal-interview/personal/<person>/user.md Company-specific facts → skills/proposal-interview/companies/<slug>/org.md Multi-person support: Default: current user If doing this for someone else (e.g., spouse), use skills/proposal-interview/personal/<their_name>/ Relationships: If a company has an owner/founder relevant to letters, record in org.md
First time with a person (onboarding): Ask 10 initial discovery questions in one session Then generate first batch of statements All subsequent sessions (updates/refinement): Ask only 3 questions per round Then generate statements from those 3 answers This keeps sessions manageable and can run via cron for continuous learning Why this matters: You can schedule this skill as a cron job to periodically learn more about the person (every week, every month, etc.). The 3-question pattern keeps token usage reasonable while continuously building your statement library.
Read this skill document to understand the workflow Create folder structure if it doesn't exist: mkdir -p skills/proposal-interview/personal/<user_name> skills/proposal-interview/companies Create profile.md with current snapshot (name, what they do, where, objectives) Explain the process to the user: "I'm going to ask you 10 onboarding questions to build your initial profile. After this first session, future updates will only ask 3 questions at a time to keep things manageable. This can even run on a schedule to continuously learn more about you." Run onboarding questions (see below) Ask the 10 initial discovery questions Draft 3-8 statements, get approval, store approved ones Future sessions: switch to 3-question rounds
CRITICAL: Always read existing files BEFORE asking questions or appending facts. This prevents duplicates and helps you incorporate existing knowledge. Read existing files IN THIS ORDER: skills/proposal-interview/personal/<user>/profile.md (READ FIRST - current context) skills/proposal-interview/personal/<user>/user.md (facts gathered - NEVER DELETE, ONLY APPEND) skills/proposal-interview/personal/<user>/statements.md (approved statements - NEVER DELETE, ONLY APPEND) skills/proposal-interview/personal/<user>/preferences.md (style guide) skills/proposal-interview/personal/<user>/coherence.md (gaps/conflicts) Ask only 3 new questions (fill gaps, drill down on threads) Append new facts to user.md (NEVER delete existing facts, always append) Draft 2-4 new statements based on new facts Get approval, append to statements.md (NEVER delete approved statements, always append)
Create company folder: skills/proposal-interview/companies/<slug>/ Ask company-specific questions: What do they do? Domain? Key people? Relationships? Why is user interested? What's the user's angle/fit? Store company facts in org.md Draft company-specific statements (e.g., "why I'm a fit for Acme Robotics") Store in company's statements.md
First-ever run for a person: Ask exactly 10 questions before generating statements After first batch: Ask 3 questions per round Generate statements Get feedback Update files Repeat If user gives very short answers: Comment briefly: "That's shorter than ideal, but I can work with it." Ask a follow-up only if critical
When starting with a new person, ask these setup questions:
"Are you creating this for yourself, or for someone else (e.g., spouse, client)?" If for someone else, create skills/proposal-interview/personal/<their_name>/
"Which platforms do you apply through? Check all that apply: Upwork / Freelance marketplaces LinkedIn (applications or outreach) Email outreach (cold/warm) Company career portals Grants / funding applications Other (specify)"
For each platform they checked, ask: "For [platform], what's your preferred style? Tone: confident / warm / direct / humble / technical / playful / formal Length: short (2-4 sentences) / medium (1-2 paragraphs) / long (3-4 paragraphs) Structure: bullets / narrative / metrics-first / story-first What does success look like? (get a reply / win a contract / advance to interview)"
"What type of work do you want more of? What's your unfair advantage?"
"Any topics, buzzwords, or sensitive info we should avoid in proposals?" After onboarding, write preferences to skills/proposal-interview/personal/<user>/preferences.md.
After onboarding, ask these 10 questions (adapt based on their work/goals):
"Where have you lived or traveled that shaped how you work? (Dubai, Seoul, remote teams, etc.) Answer in 3 bullets or a short paragraph."
"What have you learned in your work that doesn't fit neatly on a resume? What was surprisingly hard and how did you handle it?"
"What do colleagues, clients, or teammates repeatedly come to you for? What's your 'thing'?"
"Any classes, courses, or learning experiences that changed how you think or work? (Formal or self-taught—include MOOCs, bootcamps, workshops.)"
"What's a project you're proud of that shows your skills? (Side project, work project, personal tinkering, GitHub repo, portfolio piece.) Describe scope and outcome."
"Any awards, scholarships, competitions, certifications, or public recognition? (School, work, industry, community.)"
"What hobbies did you used to do a lot but don't anymore? What skills did they build? (Sports, arts, teaching, community involvement, etc.)"
"Any books, ideas, or people that shaped your approach to work or leadership? (Use sparingly, only if it produces something concrete.)"
"Tell me about a role, class, or project you found boring or frustrating. What did you learn? What do you now avoid or demand in your work?"
"If you had to pitch yourself in one sentence, what's your unfair advantage? What makes you different from 100 other people with similar skills?" After these 10 questions: Extract facts → append to user.md Draft 3-8 statements (see below) Get approval Store approved statements → statements.md Update preferences.md based on feedback
After the initial 10 questions, switch to asking 3 questions per round. How to choose the next 3 questions: Check coherence.md for gaps or conflicts that need clarification Review statements.md to see which categories are under-represented Drill down on strong threads: If user mentioned robots → ask about tools, scope, outcomes, demos If user mentioned Dubai/Korea → ask about cross-cultural teamwork, languages, client communication If user mentioned a specific company → ask domain-specific questions Question design principles: Allow short or long replies Use phrasing like: "Answer in 3 bullets or a short paragraph." If dates/metrics uncertain: "Approximate is fine." Always explain why it matters (briefly): "This helps me write more personal proposals." After each 3-question round: Extract facts → append to user.md or org.md Draft 2-4 new statements Get approval Update files
After the first 10 questions (initial run) After every 3-question round (subsequent runs)
Generate 3-8 candidate statements per round. Each statement must: Be grounded in gathered facts (no invention) Be usable in proposals/cover letters Be written in multiple variants when helpful: Upwork short pitch (1-2 lines) Standard cover letter (2-4 sentences) Technical/proof-first variant (metrics, tools, outcomes) Statement structure: --- statement: "[The actual statement text]" tags: #tag1 #tag2 #tag3 evidence: user.md line X, org.md line Y platform: Upwork short / standard cover letter / technical ---
For each statement, ask the user: "Here's a draft statement. Choose one: Approve - Store as-is Edit - Suggest changes (I'll rewrite and confirm) Reject - Don't store, move on" If user chooses Edit: Propose a rewrite Get confirmation If confirmed, store If user chooses Approve: Read existing statements.md first to avoid duplicates Append the approved statement to statements.md in the correct folder (skills/proposal-interview/personal/ or skills/proposal-interview/companies/) NEVER delete previously approved statements If user chooses Reject: Don't store (obviously, since it's rejected) Note the rejection reason in preferences.md if it reveals a style preference ⚠️ CRITICAL REMINDER: Once a statement is approved and appended to statements.md, it must NEVER be deleted. Only append new approved statements.
After each approval round, update preferences.md if you learn: Tone preferences (user prefers "confident" over "humble") Length preferences (user always wants shorter) Structure preferences (user likes bullets over narrative) Word choice (user hates "passionate", prefers "focused on")
Example: I've built robotic arms at home using Arduino and 3D-printed parts—this isn't just work for me, it's how I spend my evenings. I bring 5 years of experience with ROS, Python, and mechatronics, plus a global perspective from working with EMEA and APAC clients. • Built 3 functional robotic arms with custom inverse kinematics • MIT OCW Machine Learning graduate (applied to robotics path planning) • Available for 20-30 hrs/week, overlap with US/EU time zones
Example: I bring a global perspective from living and working across three continents—from collaborating with EMEA clients in Dubai to navigating cross-cultural teams in South Korea. I've learned to adapt communication styles and build trust across cultures, which is essential for remote, distributed work. My technical background in robotics and hands-on engineering complements my ability to work with diverse teams and deliver practical solutions.
Example: I've designed and built three functional robotic arms using Arduino, custom inverse kinematics algorithms (Python), and 3D-printed components. I completed MIT OCW's Machine Learning course and applied it to a path-planning optimization project that reduced movement time by 18%. My GitHub portfolio includes ROS packages, simulation environments (Gazebo), and documentation that's been forked 40+ times.
Example: Hi [Name], I noticed Acme Robotics is hiring for a Senior Robotics Engineer. I've been tinkering with robotic arms at home for years—building them from scratch using Arduino and 3D-printed parts—and I'd love to bring that hands-on passion to a team working on real-world automation challenges. I've worked with clients across Dubai and South Korea, so I'm comfortable with remote collaboration and adapting to different working styles. Would you be open to a quick call to discuss the role? Best, [User]
Example: My unique combination of hands-on robotics experience and cross-cultural collaboration positions me well for this initiative. Over the past four years, I have: 1. **Designed and built functional robotic systems**: I independently developed three robotic arms using Arduino microcontrollers, custom inverse kinematics algorithms (Python), and 3D-printed mechanical components. These projects demonstrate my ability to work across hardware, software, and mechanical design. 2. **Applied machine learning to robotics**: I completed MIT OpenCourseWare's Machine Learning curriculum and applied those techniques to optimize path planning for robotic arm movement, achieving an 18% reduction in task completion time. 3. **Collaborated across cultures and time zones**: My professional experience spans Dubai (EMEA clients) and South Korea (APAC teams), where I learned to navigate language barriers, cultural differences, and asynchronous communication. I speak English natively, Korean at an intermediate level, and basic Arabic. This background enables me to contribute both technical expertise and the adaptability required for distributed, international teams.
Maintain an internal checklist of what categories are covered: Geographic / cultural context Work experience beyond resume Education / formative learning Projects / proof Awards / recognition Past hobbies Books / influences Boring stories / frustrations Proposal leverage / unfair advantage Constraints / boundaries Prioritize questions that: Fill gaps in under-represented categories Drill down on strong signals (unique experiences like living abroad) Produce demonstrable proof (projects, metrics, artifacts) Align to target work (if a company/initiative is provided)
If user mentions robots: "What tools did you use? What was the scope? Any metrics or demos?" "What's sitting in your office/workshop right now? How does it map to client value?" If user mentions Dubai/Korea: "Tell me about a time cross-cultural communication was challenging. How did you adapt?" "What languages do you speak? At what level?" "How did living there change your approach to remote work or global clients?" If user mentions a specific job/company: Prioritize questions that produce domain-aligned statements Store company-specific learnings in skills/proposal-interview/companies/<slug>/
CRITICAL WORKFLOW - DO THIS EVERY TIME: BEFORE appending any facts: READ the existing user.md or org.md file to check what's already captured Check for duplicates - don't re-add facts that are already there Incorporate existing facts when generating statements After each user answer: Read existing facts first (see above) Extract discrete factual lines from the answer Append them to the correct file (NEVER delete existing content): Personal facts → skills/proposal-interview/personal/<user>/user.md Company facts → skills/proposal-interview/companies/<slug>/org.md If a fact is uncertain, ask a one-line confirmation before writing: "You mentioned living in Dubai—was that 2018-2020, or different years?" Never delete or rewrite previous lines. If a conflict is detected: Append the new info (DO NOT DELETE THE OLD INFO) Append a "possible conflict" note to coherence.md Ask a targeted clarifying question next round REMINDER: APPEND ONLY. NEVER DELETE. ALWAYS READ BEFORE WRITING.
Only write statements grounded in gathered facts If user didn't mention it, don't make it up
Don't include salary history, medical details, or private info unless user explicitly provides and approves it
Refuse politely: "I can't invent achievements, but I can help you frame what you've done more compellingly. Let's work with the real facts." Offer truthful alternatives
If something doesn't add up, note it in coherence.md and ask for clarification Never guess or fill in blanks
You can schedule this skill to run periodically (weekly, monthly) to continuously learn more about the person and build their statement library. Example cron job payload: { "kind": "agentTurn", "message": "Run the proposal-interview skill for Mike. This is a continuation session (not first run), so ask 3 new questions, gather facts, draft 2-4 statements, and get approval. Focus on filling gaps in his profile or drilling deeper on OpenClaw implementation experience." } Benefits of cron-based learning: Builds statement library over time without overwhelming the user 3-question rounds keep sessions short and token-efficient Can be scheduled during low-activity periods Gradually captures more depth and nuance Important: Always use the 3-question pattern for cron sessions, never the initial 10-question onboarding.
⚠️ CRITICAL CHECKS - READ THESE EVERY TIME: Before you start: Have I identified which person/company this is for? Have I read profile.md FIRST to get current context? Have I read existing user.md or org.md to see what facts are already captured? Have I read existing statements.md to avoid duplicating approved statements? Have I read preferences.md and coherence.md? Have I determined if this is a first run (10 questions) or continuation (3 questions)? During the interview: Am I asking questions that allow short or long replies? Am I reading existing facts BEFORE appending to avoid duplicates? Am I appending facts (NEVER deleting) to the correct file (user.md or org.md)? Am I noting conflicts in coherence.md instead of rewriting? Am I tagging statements with relevant keywords? When drafting statements: Have I read existing statements to avoid duplicates? Are all statements grounded in gathered facts? Have I written variants for relevant platforms (Upwork, cover letter, technical)? Have I asked: Approve / Edit / Reject for each statement? Have I appended approved statements (NEVER deleting old ones) to statements.md? Have I updated preferences.md based on user feedback? After the session: Have I noted any gaps or missing info in coherence.md? Have I identified categories that need more coverage for next round? Have I maintained the append-only rule (no deletions, only additions)? Did I update profile.md if any current information changed?
This skill transforms generic proposals into personalized, compelling letters by: Interviewing you to extract unique, personal-but-professional facts Drafting reusable statements grounded in those facts Refining statements through an approval loop Storing approved statements for future assembly The result: a library of high-quality, pre-approved statements you can mix and match for any proposal, tailored to the platform and company. Next steps: Run the onboarding + 10 initial questions Approve your first batch of statements Run 3-question rounds to fill gaps and drill down Build a library of 20-50 statements covering different angles (Future) Use those statements to assemble custom proposals on demand File locations: Personal: skills/proposal-interview/personal/<name>/ Companies: skills/proposal-interview/companies/<slug>/ Always append, never delete Check coherence.md for gaps
Workflow acceleration for inboxes, docs, calendars, planning, and execution loops.
Largest current source with strong distribution and engagement signals.