skills-update

name	skills-update
description	Real-time skills matrix updates. Adds, modifies, or marks stale any skill in the user's skills-matrix.md. Activates when the user mentions learning a technology, disputes a gap flagged by the job-match-scorer, or asks to update their skills profile. On gap closure, offers to re-score affected pipeline roles. Owns all writes to: identity/skills-matrix.md (xOS primitive) via brain.write().
triggers	["su","add to my skills","update my skills","update my skills matrix","mark as proficient","I'm now proficient in","I learned","I've been working with","I've been doing a lot of","I built something with","I shipped","add proficiency in","remove from my skills","that skill is stale","rescan my skills","that's not actually a gap","I do know","I have experience with"]

Skills Update — Career OS Skill

Purpose

Adds, modifies, or marks stale any skill in the user's skills-matrix.md. Handles both individual real-time updates and bulk rescan. On gap closure, offers to re-score affected pipeline roles via the job-match-scorer.

Separated from organize because organize handles bulk one-time scanning; skills-update handles the ongoing, real-time maintenance of the skills matrix.

Output Format

Always start your response with:

━━━ Career OS: Skills Update ━━━

How to Invoke

• "I learned Rust" — add a new skill to the matrix
• "add Python to my skills" — explicit add
• "mark Go as proficient" — update proficiency level
• "that's not actually a gap" — dispute a job-match-scorer-flagged gap
• "rescan my skills" — bulk GitHub repo rescan
• "remove React from my skills" / "React is stale" — mark as stale

---

DATA ARCHITECTURE

Brain API (brain-kernel >= 1.0.0)

identity/skills-matrix.md is an xOS primitive. This skill writes it via brain.write() with engine_id: "career-intelligence" — permitted because identity/skills-matrix.md is in this engine's writes_to_primitives declaration.

Inputs (what the skill reads)

Source	brain.read() path	What It Provides
Skills matrix	brain.read("identity/skills-matrix.md")	Current skills state
Match Tracker	brain.read("career-intelligence/projects/job-search/job-pipeline-match-tracker.json")	Active pipeline roles with gap flags
Pipeline	brain.read("career-intelligence/projects/job-search/job-pipeline.json")	Companies to check for gap closure impact

Outputs (what the skill writes)

Output	brain.write() path	What It Contains
Skills matrix (updated)	identity/skills-matrix.md	New/modified skill rows
Handoff entry	NEXT_SESSION_HANDOFF.md	Note on significant changes (especially gap closures)
Console output	—	Re-score offer for affected pipeline roles

Write call pattern (primitive write):

brain.write("identity/skills-matrix.md", content, {
  provenance: { who: "career-intelligence", why: "skill updated by user", source: "skills-update" },
  engine_id: "career-intelligence"
})

---

PROFICIENCY & LEARNABILITY SCALES

Proficiency

• Expert — deep production experience, can mentor others
• Proficient — solid working knowledge, production use
• Familiar — have used it, could ramp quickly
• Learning — actively acquiring, not yet production-ready
• Gap — no meaningful experience
• Stale — previously known, no longer current

Learnability

• Fast — closes in <1 month (adjacent skill, strong foundation)
• Medium — 1-3 months (new domain area)
• Slow — 3+ months (structural, requires domain depth)
• N/A — already proficient or above

---

BEHAVIOR: Add a Skill

1. Identify the correct category (Languages & Frameworks, Platforms & Infra, Domain, Leadership)
2. Propose the row: skill name, proficiency, last used, evidence, learnable
3. Show the proposed addition for confirmation — never write without approval
4. Write on approval — append to the correct table section
5. Update Last updated date in the file header

---

BEHAVIOR: Update Proficiency

1. Find the existing row in skills-matrix.md
2. Show current vs. proposed value
3. Update on confirmation
4. If proficiency improves from Gap/Familiar to Proficient/Expert:
   • Check if job-match-scorer has flagged this as a gap for any active pipeline roles
   • Surface: "You just improved [X] — this closes a gap at [Company]. Want me to re-score?"

---

BEHAVIOR: Mark Stale / Remove

1. Don't delete — mark as Last Used: {date} (stale) and Proficiency: Stale
2. Stale skills still matter for historical context but are de-weighted in scoring
3. Confirm before marking stale

---

BEHAVIOR: Bulk Rescan

When triggered by "rescan my skills":

1. Re-read GitHub repos via available tools
2. Diff against current matrix
3. Show: new skills detected, proficiency changes, stale candidates
4. Confirm before writing — present the full diff
5. Update matrix on approval

---

BEHAVIOR: Gap Dispute

When the user disputes a job-match-scorer-flagged gap ("that's not actually a gap", "I do know [X]"):

1. Find the skill in the matrix
2. If it exists at Gap/Familiar level → offer to update proficiency
3. If it doesn't exist → offer to add it
4. After update, offer to re-score the affected role

---

SCORER INTEGRATION (P14 Self-Evolution)

The job-match-scorer reads skills-matrix.md for:

1. Gap detection — is a required JD skill at Proficient or above?
2. Learnability — if a gap, how fast can it be closed?

When skills-update runs, check active pipeline roles in job-pipeline-match-tracker.json:

"You just added [X] — this closes a gap at [Company]. Want me to re-score that role?"

This closes the self-evolution loop: skill improves → score improves → pipeline updates.

---

INTEGRATION POINTS

Skill	Relationship
job-match-scorer	Primary consumer — skills-update feeds job-match-scorer accuracy. Gap closure triggers re-score offer.
resume-engine	Uses skills matrix to select which technologies to highlight per track.
organize	Bulk skills scan is a one-time operation; skills-update handles real-time changes.
interview-prep	May surface skills gaps when mapping stories to JD requirements.
mission-control	Memory health section could surface "skills matrix last updated X days ago".

---

MULTI-AGENT SAFETY (P15)

1. brain.read("identity/skills-matrix.md") immediately before any write (kernel pull ensures latest)
2. Surgical edits — append rows or update specific cells, never rewrite tables
3. Update Last updated date on every write
4. Log significant changes to NEXT_SESSION_HANDOFF.md (especially gap closures)

---

EDGE CASES

• Duplicate skill: warn "X already exists at [level] — want to update proficiency instead?"
• skills-matrix.md doesn't exist: write via brain.write("identity/skills-matrix.md", ...) with header structure and category sections
• No active pipeline roles: gap closure re-score offer is skipped gracefully
• User disputes but evidence is ambiguous: ask for evidence ("What project did you use X on?") before updating
• Bulk rescan with no GitHub access: surface the limitation, offer manual add instead