$ vibexp profile @dvk31 --verified v18
fetching profile … verified snapshot v18 · sha256:8ed8cf80297a… · stats updated 11h ago · next AI analysis 7d from now
// BUILDER · #1 OF 1 ALL-TIME
“Full operator enriched preview: the builder profile and artifact scorecards now use one evidence model.”
profile · vibexp verified v18
grade91/100 [Exceptional] · rank #1 of 1 all-timecompositionsession craft 92 × 65% · artifact quality 88 × 35%levelPlatform Shaper (proposed · preview) · capability 92/100 · conf 90% confidenceroleAI Platform OperatorspecialtiesTooling · Security · Distributed systems · Scalabilityidentity@dvk31 · snapshot ✓ verified
$ signature-moves
what this builder repeatedly does well — distilled from scored sessions
├─ designs reusable agent context systems and local tool surfaces
├─ turns platform boundaries into practical operating rules for agents
├─ builds quality gates that agents can repeatedly execute
├─ sustains high volume work across multiple coding agents
╰─ ships product infrastructure that compounds across later work
$ scores --by-dimension
confidence shown, not hidden · grounded in redacted session excerpts
architecture leads — Operator manuals and local tools show a platform architecture practice: boundaries, capability checks, shared substrate, and reusable context surfaces are designed before implementation.
$ evidence --redacted | head -8
architecture33f8b68ccb2722e6117446de
Guidance separates platform authority, runtime execution, gateway routing, and builder work before implementation.
└ Shows durable architecture boundaries for agents.
planning33f8b68ccb2722e6117446de
Skill routing breaks work into platform, identity, gateway, runtime, builder, and quality tracks.
└ Shows explicit task partitioning before edits.
steering33f8b68ccb2722e6117446de
Operating rules tell agents which contracts and quality checks must guide each task.
└ Shows steering through durable rules.
debugging & rigor33f8b68ccb2722e6117446de
Health endpoints, schema checks, redaction scans, and verifier gates are treated as first class failure signals.
└ Shows rigorous debugging signals.
code quality33f8b68ccb2722e6117446de
Contract pinning and redaction tests catch drift between client and server behavior.
└ Shows review by invariant, not just diff reading.
execution33f8b68ccb2722e6117446de
The profile shows sustained work across many agent sessions and multiple local tools.
└ Shows high throughput under agent leverage.
product instinct33f8b68ccb2722e6117446de
Profiles, leaderboards, living refresh, shipped records, and installer flow form one coherent product surface.
└ Shows product integration beyond a single feature.
communication33f8b68ccb2722e6117446de
Agent guidance is written as concise operating rules that can be followed by different tools.
└ Shows communication as executable context.
$ shipped --builder-approved
outcomes this builder explicitly approved for publication · last detected jun 11, 2026
├─ mar 2026 revamp marketing homepage and footer · deployed
├─ apr 2026 add runtime cursor sdk probe gate · deployed
├─ may 2026 developer site + docs · deployed
├─ jun 2026 prepare schema compiler 0.1.6 release · deployed
├─ jun 2026 repo arch profile block - schema and skill assembly with multi-repo upload selection · git backed
╰─ jun 2026 render repo architecture scorecards on profile detail · git backed
named items are builder-approved; private local signals stay local
$ repos --scorecards --self-reported
open repo rubric v1.0.0 · generated locally · not yet server re-verified
├─ repos scanned ·············· 12 · python + typescript
├─ mean overall ·············· 87.9/100 · 10.0/10 dims scored
├─ python · very large repo ×3 · Exceptional 92/100 · coverage 10/10 · architecture 9.8 · maintainability 9.4 · testing 9.3
├─ python · very large repo ×2 · Exceptional 90/100 · coverage 10/10 · architecture 9.7 · maintainability 9.2 · testing 9.1
├─ typescript · medium repo · Strong 87/100 · coverage 10/10 · architecture 9.4 · maintainability 8.9 · testing 8.8
├─ typescript · medium repo · Strong 86/100 · coverage 10/10 · architecture 9.3 · maintainability 8.8 · testing 8.7
├─ python · small repo · Strong 86/100 · coverage 10/10 · architecture 9.3 · maintainability 8.8 · testing 8.7
├─ typescript · tiny repo · Strong 85/100 · coverage 10/10 · architecture 9.2 · maintainability 8.7 · testing 8.6
├─ python · tiny repo · Strong 85/100 · coverage 10/10 · architecture 9.2 · maintainability 8.7 · testing 8.6
╰─ typescript · medium repo ×2 · Strong 85/100 · coverage 10/10 · architecture 9.2 · maintainability 8.7 · testing 8.6
repos identified by language and size band only; names never leave the machine
$ capability --operating-level
Builders who design and standardize systems that others build on.
shows reusable context management rather than one off prompting.
local guidance routes agents through platform, runtime, gateway, quality, and builder rules before edits.
next level: External adoption, team workflows, and production telemetry unlock the next platform tier.
$ output --by-period
metrictodayweekmonthyear
vibe loc117,43888,023578,763
agent-hours23.2h128h346h1,231h
calendar hours6.8h34.5h106h443h
provider tokens393.2M8.1B14.3B43.8B
cursor provider tokens unavailable · agent commit attribution inferred · bulk import loc excluded
$ today --local-day
├─ agent time ·············· 6.3h
├─ active time ·············· 1.8h
├─ sessions ·············· 27
╰─ prompts ·············· 730
output mix other 1
codex cli 1.8h · claude code 3.9h · cursor 0.6h · last activity 11h ago
$ performance --insights
├─ vibe agent-hours ·············· 1,231h · Pilot-style AI build time
├─ active calendar ·············· 443h · Deduped wall-clock time
├─ provider tokens ·············· 43.8B · Tracked provider counters
├─ vibe loc ·············· 578.8K · Smart local output count
├─ shipped outcomes ·············· 6 · Builder-approved for public display
├─ longest streak ·············· 24 · consecutive build days
├─ edit precision ·············· NA · edits accepted first pass
╰─ ai analysis ·············· Weekly · Next 7d from now
$ identity --stats
who they are as a builder — tenure, loyalty, pacing, streaks
├─ building with AI ·············· 85 days · since March 2026
├─ primary harness ·············· Cursor · 94% of active days · all 3 tools on 11 days
├─ session shape ·············· Sprinter · median session 7 min
├─ current streak ·············· 17 days · consecutive active build days
╰─ longest streak ·············· 24 days · longest run of consecutive build days
$ discipline --ship
├─ test before ship ·············· 96% · commits had a test run first
├─ plan to ship ·············· 92% · planning sessions shipped
├─ blast radius ·············· 9 · files in a median commit · p90 35
├─ edit precision ·············· NA · edits accepted on the first pass — needs ≥30 sampled edits (n=22)
╰─ rework 48h ·············· 41% · of shipped lines reworked within 48h
$ efficiency
├─ context from cache ·············· 54% · prompt context reused from cache
├─ best 7-day stretch ·············· 156.4K · 73 commits
├─ churn ratio ·············· 0.20 · recent deletions vs additions
╰─ token efficiency ·············· 12 · authored LOC per million tokens
$ orchestration
╰─ deepest delegation ·············· 1 · levels of subagents
coordinates codex cli, claude code, cursor and subagent workflows
$ models --share-of-agent-hours
share of total agent-hours per harness
$ tools --coverage
toolsessionsagent-hrstokens
claude code427220h8.5B
codex cli474541h35.3B
cursor483470htokens unavailable
public coverage only · unavailable provider token sources stay explicit
$ rhythm --local-clock
peak hour 2 PM
biggest push: 52K lines in 32 commits — a Friday
├─ merged prs (30d) ·············· 36 · merge-commit heuristic
╰─ night-owl share ·············· 3% · 22:00–02:00 commits
$ growth-edges
make outcome evidence more public and builder approved
promote operator manual evidence into the real scoring contract
keep raising ordinary repo hygiene to match the operating model
$ verify --provenance
how this snapshot was produced and protected
├─ snapshot ·············· v18 · sha256:8ed8cf80297a…
├─ generated ·············· Jun 11, 2026 · rubric 1.1.0 · schema 1.3.0
├─ evidence base ·············· 20 episodes scored · 40 sessions · 20 active days
├─ analysis depth ·············· 5.8M of 43.6B tracked tokens read by the scorer
╰─ redaction ·············· fail-closed · 0 fields dropped · secret + code-path scrubbers applied
profile rendered in public terminal mode
every claim carries public profile data · private local signals stay private
$