| name | profile-intelligence |
| description | Gather [OWNER]'s personal documents (academic essays, speeches, startup archives, assessments), convert them to Markdown, triage for signal, and synthesize a comprehensive Founder Profile Intelligence MOC plus any downstream profile docs (YC founder profile, bios, about pages). Use when [OWNER] says "profile intelligence", "founder intelligence", "build my founder profile", "synthesize who I am", or needs a personal/founder profile assembled from his real history. |
Profile Intelligence
Turn [OWNER]'s scattered personal record — academic work, competition speeches, startup archives, personality assessments, and the CKIS knowledge base — into a single coherent map of who he is as a founder and person, then power any downstream profile artifact (YC founder profile, investor bio, about page, personal-brand narrative) from that map. This is the skill that makes "tell them who I am" produce truth from evidence instead of generic founder fluff.
Codifies the workflow first executed 2026-05-25 to build [[05-knowledge/permanent-notes/MOC-Profile-Intelligence]] and [[03-projects/[your-project]/yc-application/YC-S26-founder-profile]].
Workflow
- Define the target. Confirm what's being produced: just the intelligence MOC, or the MOC + a specific downstream doc (YC profile, bio, etc.). If a downstream doc, get its exact field list / form structure up front.
- Gather sources. Collect from both the vault and external document folders:
- Vault:
05-knowledge/permanent-notes/ (identity, philosophy, alter-ego, startup-ideas-backlog), 05-knowledge/patterns/, 05-knowledge/maps-of-content/, 11-archive/ (every file — past projects/applications), 01-systems/ckis/01-ckis-user-profile-and-operating-context.md, 00-inbox/_PROFILE.md/_MEMORY.md.
- External (read-only):
~/Documents/Academic/ (essays, vocational work, IKIGAI), ~/Documents/Startups/Archieve/ (Fletex, BusNET, We Solve, Movi, Hidro+, ePOS, etc.).
- Convert to MD if needed. Non-MD sources (
.docx, .pdf, .txt, audio transcripts) get converted/read. Stage conversions in 00-inbox/tmp-*/convert-queue/ — never edit originals. Use the convert-to-md skill for batch conversion. PDFs: read directly with the Read tool (use pages: ranges for >10-page PDFs). Audio: read the existing (transcript).txt siblings.
- Triage for signal. Not every document matters. For each source, extract: founder-DNA evidence, idea-evolution data points, thinking-framework instances, things built, communication/oratory record, values/mission/anchors, market insights. Discard administrative noise. Preserve real facts verbatim — grades, dates, placements, metrics. Never inflate or invent.
- Synthesize the MOC. Build/update
05-knowledge/permanent-notes/MOC-Profile-Intelligence.md with these sections (adapt as evidence dictates):
- Founder DNA (repeating patterns across all sources)
- Idea Evolution (chronological table + the connecting thread)
- Thinking Frameworks (link each to its permanent note)
- Building Portfolio (hardware / software / content / services)
- Oratory & Communication
- Personal Anchors (IKIGAI, alter ego, philosophy, disciplines)
- Market Insights
- The Pitch Profile (the fundable/synthesized view)
- Vault Cross-References (every node, with
[[wikilinks]])
- Produce downstream doc(s). If a YC founder profile / bio / about page was requested, fill every field in [OWNER]'s voice (direct, honest, concrete, no buzzwords), sourced from the MOC. Frontmatter
related: links back to the MOC.
- Wire the graph. Add bidirectional
[[wikilinks]] between the MOC, the downstream doc, and every source permanent note referenced. Update modified on any existing note touched.
- Report inline. Summarize the key founder-intelligence patterns discovered (3-5 sentences) and list every file created/updated with absolute paths.
Output format
- Primary artifact:
05-knowledge/permanent-notes/MOC-Profile-Intelligence.md (the 9-section map above).
- Downstream artifacts (as requested): profile docs in the relevant project folder (e.g.
03-projects/[your-project]/yc-application/).
- Inline report: patterns discovered + file manifest with absolute paths.
Rules
- Read before edit. Every file read before any modification. Never edit a note you haven't read this session.
- Preserve originals. Never edit source documents in
~/Documents/Academic/ or ~/Documents/Startups/. Convert into the vault's convert-queue; treat originals as read-only.
- No deletion. Move to
11-archive/ if anything must be retired. Never rm.
- Facts are sacred. Add insight and connection, never change the underlying facts (grades, dates, placements, financials, who-did-what). When uncertain about a fact, flag it rather than smooth it over.
- Preserve language. Read Spanish sources in Spanish; do not translate vault notes. Downstream English artifacts (YC, US-facing) are written in English, but the underlying Spanish sources stay Spanish.
- [OWNER]'s voice in profile prose. First person, direct, confident-not-arrogant, concrete, no buzzwords. The voice reference is
[[01-systems/ckis/01-ckis-user-profile-and-operating-context]] §Work Style.
- Synthesis is prose, not citation dumps. The "Founder DNA" and "Pitch Profile" sections must read as connected insight, not bullet lists of sources.
- Wikilink liberally. The MOC's value is its connection density — link every framework, project, and archive node.
- Honesty over polish. If a decision is made-but-not-executed (e.g. scholarship resignation), say exactly that. If a claim can't be sourced, don't make it.
Gotchas
- PDFs over 10 pages fail a single Read — use the
pages: parameter in ranges, or it errors out.
- IKIGAI / vocational PDFs live deep in
~/Documents/Academic/Programa Oportunidades FGK/... with ~$-prefixed lock files (LibreOffice temp files) that are NOT the real document — match the real .pdf/.docx, ignore ~$ siblings.
- Audio sources (e.g. Hidro+, Fletex) already have
(transcript).txt siblings — read those, don't try to process audio.
- The freight transcript (Fletex) is [OWNER]'s father, a truck driver, describing the problem — attribute the idea origin correctly ([OWNER] saw the gap from his father's experience), don't mis-read it as [OWNER]'s own job.
- ePOS was discarded — but the discard reason is the signal, not a failure. He rejected the static/generic POS model and the thesis re-emerged in [YOUR_PROJECT]. Frame discards as maturity, not abandonment.
- Don't treat school essays as "just homework." The NASA/YA/IKIGAI/admission essays are deliberate identity documents (same lesson as the Jackson Steele extraction in
[[05-knowledge/patterns/chatgpt-knowledge-extraction-2026-04]] §Gap 2). Mine them for self-model, not just facts.
- Personality assessments (ENTJ-A, FODA) are self-reported — present as self-assessment, not clinical fact.
- Grades are on a 0–10 scale (Salvadoran system: 9.8/10, 9.5/10) — don't convert to GPA or mislabel as percentages.
- The idea-evolution thread is the highest-value output for fundraising — always trace the convergence (We Solve → HealthFlow → ePOS → Movi → Tourdy → [YOUR_PROJECT]) rather than listing ventures in isolation. The narrative "this is a 3-year thesis, not a fresh idea" is the founder-market-fit proof.
Example invocation
[OWNER]: build my founder profile for [program/investor]
→ Confirm target: intelligence MOC + the program's exact field list
→ Gather: vault permanent-notes + 11-archive + Academic essays + Startups archive
→ Convert/read non-MD sources (PDFs in page-ranges, .docx via convert-to-md, audio transcripts)
→ Triage each for founder-DNA / idea-evolution / portfolio / oratory / anchors / market
→ Build MOC-Profile-Intelligence (9 sections, heavy wikilinks)
→ Fill the program's profile fields in [OWNER]'s voice, sourced from the MOC
→ Wire bidirectional links; update modified dates
→ Report patterns discovered + file manifest (absolute paths)