Progress - YakDriver/fardvag GitHub Wiki
Promotion Master Context — SuccessFactors Goals + BadgeME (2026)
Mission: support a promotion. Two workstreams feed it: (A) SuccessFactors mid-year goals and (B) BadgeME — Developer Profession, Level 3 (Expert). Both draw from the same evidence (see §1). Applicant: Dirk A. / GitHub YakDriver — Terraform AWS Provider, Go, AWS; public OSS maintainer.
1. Shared evidence library (canonical — cite from here for BOTH workstreams)
Business-value facts (point-in-time, June/July 2026 — cite a dashboard/registry source):
- Most-used Terraform provider: 6.8B all-time downloads; ~40M monthly RUM = 62.49% of all provider usage; 65,843 organizations; RUM +2.63% MoM, +13.74% YTD.
- Public AWS partnership depends on the provider; material to HashiCorp/IBM value.
- RUM = Registry usage metric (HCP). Use "tens of millions" if 40M can't be sourced live.
Delivery (H1 2026): 113 PRs merged, 233 PRs reviewed (reviewed-by:YakDriver); AWSCC provider releases v1.70–v1.90.
- New service
aws_s3files(file system, mount target, access point, policy, sync, data sources, sweepers). aws_wafv2_web_acl_rule(#46682) — solved a Terraform Core dependency-graph (detach-before-delete) limitation; closed 7 issues (#43739, #17601, #28331, #41069, #25669, #36856, #45858).- Resource Identity + List Resource rollout: launch_template #47540, lambda_layer_version #47496, appautoscaling_target #48449, dynamodb_table_item #48520 (closed 5 long-open: #30720/#47218/#8923/#9664/#6446), lb_target_group_attachment #46646, route #46370, SG ingress/egress #46367/#46368, route_table #46337, ecr #46344.
- glue in-place VIRTUAL_VIEW update preserving Lake Formation grants #48532.
Ownership / quality:
- Owned a self-introduced regression: AutoFlex #46741 → #46778 + guard test.
- Regressions/blast-radius fixes: secretsmanager #48318; autoscaling #46452; tags interceptor #48008 (all SDKv2 tagged resources); honor-region-on-import, 32 resources, #47043; v5→v6 upgrade tests #46553; panics route53 #47038 / cloudfront #46982 / networkmanager #46160; s3_bucket-vs-standalone #47962.
Support & bug (Goal 2): hero rotations — 19 customer issues resolved, each within the 7-day rotation window (well under the ≤3-week MTTR target). Elevance (enterprise customer) protected from the Amazon Pinpoint EOL (Oct 30, 2026) via the apicall API-call observation framework (TF-2087 / #48031) + Pinpoint EOL plan (TF-2088, #48012). ElastiCache #46670/#48462/#46454.
Automation & AI (Goal 3):
- Copilot review instructions (TF-2107) — authored general + path-specific; tracked a week at 78% actionable.
- swissshepherd doc linter (#47921, #48357; retired tfproviderdocs); makelign (#48104); apicall (#48031).
- CI speedups #48559: embedded lint ~25m→~2m; Semgrep ~13m→~5m. (verified vs PR table)
- repo-wide
terraform fmt#46339; Go test-naming #46371; Kiro AI pioneering.
Skills growth (Goal 4): 23/40 "Your Learning" hours (secure coding, Automation pillar, AI) + self-study; required IBM learning up to date (nothing overdue); secure-coding fix #48463 (HTTP body masking); new constructs (Resource Identity, List, write-only args, ephemeral).
Leadership / mentorship (Goal 5): intern Bella R. — schema-drift detection between AWS and the provider; co-authored RFC TF-2111 (Schema Coverage Analysis); with team since late May 2026; ongoing reviews/pairing. RFCs authored: TF-2107, TF-2087, TF-2088, TF-2111, TF-1124 (Top 10 Actions — all implemented) + 2026 Intern-Ideas memo. HashiConf 2025 talk with Manu C. (AWS) — "Making AI Work for You — Terraform Engineer Blueprint."
Growth behaviors (Goal 6): #46212 — challenged "AutoFlex can't do this" with a working refactor (verbatim: "three kinds of lies…").
De-jargon glossary (badge reviewers won't know these — always gloss once): the provider · Terraform · Plugin Framework / SDKv2 · RUM · AutoFlex · Resource Identity / List resources · hero rotation · swissshepherd · makelign · apicall · Kiro.
2. BadgeME — scoring rules (the gate)
Level 3 Expert requires ALL of: every Core at Expert (min); every Common at Foundation (min); ≥ 46 total points.
| Level | Pts | One-line test |
|---|---|---|
| Entry | 0 | academic only, needs supervision |
| Foundation | 1 | knows it, unsupervised but needs mentoring |
| Experienced | 2 | repeated success across environments, supervises others |
| Expert | 4 | recognized authority, handles complex/novel independently, mentors |
| Thought Leader | 8 | SME inside/outside IBM, advances the state-of-the-art |
KEY INSIGHT: floors are minimums; points accrue at the level you can evidence. A Common proven at Expert scores 4 (or TL = 8), not 1.
Point math (aligned: 6 Core, 7 Common)
- Core @ Expert floor: 6 × 4 = 24
- Common @ Foundation floor: 7 × 1 = 7
- Floor = 31 → gap to 46 = 15
Fast path to 46 (all honest)
- Building with AI (Common) → Thought Leader (8) — evidence: Copilot instructions/78%, swissshepherd, Kiro, HashiConf. Biggest lever.
- Other Commons at true level — Programming, CI, Quality, Change Mgmt, Problem Determination plausibly Expert (4).
- Core Expert→TL (+4) where defensible: Technical Leadership, Communication.
- Specialist depth: Automation (Expert/TL), Quality & Test Eng, Architecture.
Core 24 + Building-with-AI 8 + a few Commons at Expert already exceeds 46 — scale to honest levels, don't inflate.
3. BadgeME — skill inventory (target level · points · fit · evidence)
Legend: T = target level to claim; Pts at that level. Descriptor detail in badgeme-skills-reference.md.
Core (6) — must all reach Expert
| Skill | T | Pts | Evidence anchor |
|---|---|---|---|
| Technical Leadership | Expert→TL | 4–8 | Mentoring Bella; 233 reviews; RFCs; adopted tooling; HashiConf |
| Communication | Expert→TL | 4–8 | HashiConf w/ AWS; RFC authorship; exec-facing clarity |
| Problem Solving | Expert | 4 | wafv2 dependency-graph (#46682); root-cause writeups (#48462/#48318) |
| Intellectual Capital / Innovation | Expert | 4 | Reusable assets: swissshepherd, makelign, apicall, Copilot instructions |
| Technical Risk Management | Expert | 4 | Pinpoint EOL risk (Elevance); tech-debt vs delivery; region-import blast radius |
| Agile & Design Thinking | Expert | 4 | RFC-first iterative delivery; short-cycle shipping |
Common (7) — must all reach Foundation; score higher where true
| Skill | T | Pts | Evidence anchor |
|---|---|---|---|
| Building with AI | TL | 8 | Copilot instructions (78%), swissshepherd, Kiro, HashiConf (pioneer + measure + OSS/talks) |
| Programming Language Mastery | Expert | 4 | deep Go; AutoFlex/Smithy internals |
| Continuous Integration | Expert | 4 | CI speedups 25m→2m (#48559); 233 reviews; pipeline ownership |
| Quality Design & Maintainable Dev | Expert | 4 | doc-linting, fmt (#46339), guard tests, tech-debt cleanup |
| Problem Determination | Expert | 4 | cross-team debugging; subtle-defect root cause (#46452) |
| Change Mgmt & Version Control | Expert | 4 | branch strategy, hooks, repo-wide policy |
| Abstractions, DS & Algorithms | Experienced→Expert | 2–4 | design-pattern use across resources |
Specialist (pick to demonstrate; 23 available)
| Skill | Fit | T | Pts | Evidence anchor |
|---|---|---|---|---|
| Automation | High (Terraform = the doc's example) | Expert→TL | 4–8 | the provider (IaC), swissshepherd, makelign, apicall, generators |
| Quality & Test Engineering | High | Expert | 4 | acceptance-test discipline, guard tests, doc-lint gating |
| High-Level Architecture | High | Expert | 4 | s3files service; wafv2 dependency design; RFCs |
| Continuous Delivery | Medium | Experienced/Expert | 2–4 | release/pipeline (provider is a library — frame accordingly) |
| Secure Engineering | Medium | Experienced | 2 | body-masking #48463; secure-coding training |
| Security & Compliance | Situational | Foundation/Exp | 1–2 | not core focus |
| Performance Engineering | Situational | — | — | only if evidenced |
| Operating Systems | Cheap point (no TL level) | Foundation/Exp | 1–2 | shell/make/CLI; cross-platform Go; sweepers |
| Networking | Cheap point | Foundation | 1 | AWS VPC/SG/route_table/networkmanager; TLS/HTTP |
| Unusable (software=TBD): Data Science & Analytics, Service Engineering, Modelling/Sim, Storage | — | — | — | no software criteria |
N/A: Artificial Intelligence (=ml.md, model-building), Cloud & Container Orch., Program Mgmt (HW), Reliability (HW), Semiconductor/Physical/EDA/Mechatronics/SPC/Failure (HW) |
— | — | — | wrong domain — ⚠ verify if ml.md updated |
4. BadgeME — application form (structure + evidence rules)
Web form: dropdowns + text boxes. The live form is source of truth. Full field detail in badgeme-application-structure.md.
Sections: 3× Project Profile (name · dates · Business opportunity · Your Developer Contribution · Your Impact · Lessons Learned · Core/Common/Specialist multi-selects) → GiveBack (evidence for ALL 3: Guilds & communities · Mentoring & coaching · Stretch assignments) → Innovation (≥2 of: Others · Patents · Reusable Assets + 2 links) → Growing Your Skills (YL records; breadth+depth) → Eminence & Engagement (⚠ header says ≥3, dropdown says ≥2 — supply 3: Client activities · Conferences · Open source · Other · Publications/blogs) → Technical Leadership → Industry → Awards/Recognition → References (name/relationship/email).
Dropdown sets: Core (6) = Agile & Design Thinking · Communication · Intellectual Capital/Innovation · Problem Solving · Technical Leadership · Technical Risk Management. Common (7) = Abstractions · Building with AI · Change Mgmt · CI · Problem Determination · Programming Language Mastery · Quality Design/Maintainable. Specialist (23) = see badgeme-application-structure.md §2.
Evidence rules (apply to every entry):
PROBLEM — what/why it mattered (context)
CONTRIBUTION — what YOU specifically did
OUTCOME — measured result (numbers, time, users, scope)
LEARNING — (optional) what you learned / would do differently
- Map each entry to a named skill + level + points.
- De-jargon every local term (§1 glossary). Skills-based, not knowledge-based: show demonstrated behavior + outcome, never "I know X."
- Lead with context, end with quantified impact. "I built a tool" scores low.
Content-reuse map (form section → evidence):
- Project Profile 1 → s3files OR wafv2_web_acl_rule (#46682): architecture + problem-solving + quality.
- Project Profile 2 → Pinpoint EOL / Elevance (TF-2087/2088, #48031): full life-cycle + risk + partner.
- Project Profile 3 → AI/automation program (TF-2107, swissshepherd, makelign, CI #48559): automation + quality.
- GiveBack → mentoring Bella (TF-2111) + 233 reviews; OSS community; stretch = leading Copilot-review initiative.
- Innovation → Reusable Assets (swissshepherd/makelign/apicall/Copilot instructions) + git links.
- Growing Skills → 23h YL + self-study + new constructs.
- Eminence (3) → Open source (public maintainer, 6.8B downloads) · Conferences (HashiConf w/ AWS) · Publications/blogs (gap).
- Technical Leadership → mentoring, RFCs, 233 reviews, adopted tooling.
- Industry → OSS/standards influence; AWS partnership; Terraform Core dep-graph (#46682).
- Awards → gap — identify.
- References → Bella R.; Manu C. (AWS); a reviewer/manager.
5. SuccessFactors goals — status
All 6 progress reports drafted, each ≤1000 chars, business-value-forward (see h1-goalN-progress.txt).
| # | Category | Title | Primary evidence | Status |
|---|---|---|---|---|
| 1 | Business Outcomes | Provider quality & delivery | 113 PRs/233 reviews, s3files, #46682, ownership (#46741→#46778), 62% RUM share | ✅ drafted (999) |
| 2 | Business Outcomes | Support & bug resolution | 19 hero-rotation issues ≤7d, Elevance/Pinpoint, blast-radius fixes | ✅ drafted (963) |
| 3 | Business Outcomes | Automation & AI productivity | Copilot 78%, swissshepherd, makelign, apicall, CI 25m→2m | ✅ drafted (989) |
| 4 | Skills | Skills development & sharing | 23/40 YL hrs, applied AI/secure coding, 233 reviews | ✅ drafted (935) |
| 5 | Behaviors | Technical leadership & mentorship | Bella/TF-2111, RFCs, HashiConf, OSS stewardship | ✅ drafted (910) |
| 6 | Behaviors | IBM Growth Behaviors | growth (#46741→#46778), team (hero/reviews), trusted (Elevance), courageous (#46212) | ✅ drafted (1000) |
Goals ↔ Badge reuse: same evidence, inverted framing. Goals = terse, PR-cited, business-value. Badge = expanded, de-jargoned, skill+level framed (problem→contribution→outcome). Do not copy verbatim between them.
6. Open items / gaps (need from applicant)
- Verify
ml.md/ Specialist "Artificial Intelligence" — local copy still = model-building; confirm live text before using. - Confirm Common-above-Foundation adds points (needed for Building-with-AI @ TL = 8) — Competence page.
- Confirm any cap on Specialist skills counting toward 46.
- Gaps to source: a publication/blog (Eminence 3rd); awards/recognition (internal/external); an explicit stretch-assignment example; any patents/IP (else lean Reusable Assets).
- Backbone figures (6.8B / 40M / 62%) — keep a citable source; they're point-in-time.
- Fix the wiki "Progress" page — dedup to this file's structure and correct the stale "6 Common / Building-with-AI omitted" text to "7 Common, included."